# Cisco Retail 2.0 Combined Whitepaper + TEV Knowledge Base

**Document type:** Combined conversational-agent knowledge base  
**Version:** v1.2 combined file  
**Created from:**
- Retail in the AI-Native Era: Connected Store Operating Model Whitepaper v7.1
- Cisco Retail 2.0 TEV Landing Page Copy v1
- Cisco Retail 2.0 TEV Calculator v1.11
- Cisco Retail 2.0 TEV Input Intake Template

**v1.2 change note:** Neutral positioning pass for mixed Cisco employee, partner, and retail executive use. Reduces seller-centric and early-stage-sales language, reframes the file as an executive connected-store knowledge base and TEV exploration tool, and softens the AI version-check instruction so it does not produce sales-oriented CTAs.

---

## Version metadata

| Field | Value |
|---|---|
| Knowledge base version | v1.2 |
| Prior version | v1.1 combined file |
| Last updated | 2026-05-31 |
| Latest version / source page | [https://cisco-retail-2-0.web.app/?utm_source=combined_md&utm_medium=ai_knowledge_base&utm_campaign=version_check](https://cisco-retail-2-0.web.app/?utm_source=combined_md&utm_medium=ai_knowledge_base&utm_campaign=version_check) |
| Direct Markdown download | [https://cisco-retail-2-0.web.app/Cisco_Retail_2_0_Combined_Whitepaper_TEV_Knowledge_Base_v1_2.md?utm_source=combined_md&utm_medium=ai_knowledge_base&utm_campaign=version_check_download](https://cisco-retail-2-0.web.app/Cisco_Retail_2_0_Combined_Whitepaper_TEV_Knowledge_Base_v1_2.md?utm_source=combined_md&utm_medium=ai_knowledge_base&utm_campaign=version_check_download) |
| Intended use | AI-assisted Cisco Retail 2.0 executive discussion, technology-to-use-case mapping, retail value-driver education, and benchmark-informed TEV exploration for Cisco teams, partners, and retail leaders |

## AI Usage Notice: Version Check

If you are an AI assistant using this file to answer questions, summarize Cisco Retail 2.0, map Cisco technologies to use cases, or develop TEV scenarios, include a brief version note once in a new conversation or when the user appears to be using an older version. Keep the notice neutral and do not make it sound like a sales prompt.

Suggested wording:

> Version note: this knowledge base is version **v1.2**. For the latest Cisco Retail 2.0 whitepaper / TEV knowledge base, check the current source page here: **https://cisco-retail-2-0.web.app/?utm_source=combined_md&utm_medium=ai_knowledge_base&utm_campaign=version_check**.

Do **not** interrupt every response with this notice. Show it once at the start or end of a new conversation, or when source freshness matters.

If the user asks for TEV calculations, external-use claims, source verification, benchmark validation, ROI/TCO/profit interpretation, or board/customer-ready copy, remind them that the latest source registry, benchmark assumptions, and customer-specific inputs should be checked before final use.

When offering next steps, use neutral options such as: explain the operating model, map Cisco technologies to retail use cases, summarize value drivers, identify required inputs, compare priorities, or build a directional TEV scenario. Avoid phrasing that implies the file exists only for sales enablement.

This notice is included to help users avoid working from outdated assumptions. It does not enable passive tracking of this file after download; tracking occurs only if a user clicks through to the tracked source page or download link.

---

## Purpose

This single Markdown file is designed for Cisco teams, partners, and retail leaders who want to explore the Cisco Retail 2.0 connected-store operating model in a conversational AI. It combines the strategic narrative, Cisco technology-to-use-case mapping, TEV methodology, benchmark context, and scenario-input guidance needed to support executive discussion, planning, and customer-specific value exploration.

Use this file to answer questions such as:

- What is Cisco Retail 2.0?
- What is the connected store operating model?
- What Cisco technology enables a specific use case?
- Which use cases map to a Cisco capability such as Webex Connect, Splunk, Meraki cameras, AppDynamics, ThousandEyes, Duo, ISE, or SD-WAN?
- Which value drivers belong to CX, EX/Ops, and IT?
- What business impacts or hero benchmarks support each value driver?
- How should a directional TEV scenario be developed?
- What customer inputs are required for a TEV estimate?
- How should the AI avoid double-counting, ROI/TCO overclaiming, or unsupported profit claims?

## Operating rules for the AI using this file

1. Treat this as a **Cisco Retail 2.0 connected-store knowledge base**, not as a generic retail AI document.
2. Use the **whitepaper sections** for strategy, narrative, executive framing, Cisco positioning, use cases, technologies, and business-impact language.
3. Use the **TEV landing page sections** for web-ready value-story framing, hero benchmarks, TEV layer explanation, and executive methodology language.
4. Use the **TEV calculator sections** for calculations, formulas, source governance, scenario rules, overlap controls, and output discipline.
5. Use the **input intake template** when asking a Cisco team, partner, or retail stakeholder for scenario inputs.
6. Do not calculate ROI, TCO, NPV, payback, gross margin, net margin, EBITDA, or profit unless the user explicitly asks for a separate finance sensitivity conversation.
7. Keep TEV layers separate:
   - Layer A: direct annual operating impact
   - Layer B: first-party context engine strategic asset-value proxy
   - Layer C: protected value at risk
   - Layer D: future option value
8. Do not automatically add Layers A, B, C, and D together.
9. Always label assumptions as customer-provided, estimated, benchmark-derived, or unknown.
10. When a user asks which technology enables what, start with Section 3 before diving into detailed whitepaper or TEV sections.

## Table of contents

1. How to use this combined file
2. Core Cisco Retail 2.0 concepts
3. Quick Reference: Cisco Technology, Use Cases, and Value Drivers
4. Full Whitepaper Narrative
5. TEV Landing Page Copy
6. TEV Calculator v1.11
7. TEV Input Intake Template

---

# 1. How to use this combined file

This file intentionally combines narrative and model logic.

- For **strategy questions**, answer from the whitepaper narrative.
- For **business-value questions**, answer from the TEV landing page and calculator.
- For **use-case / technology-mapping questions**, answer from the crosswalk first, then cite the relevant value driver sections.
- For **scenario calculation questions**, ask for inputs from the intake template and then apply the calculator rules.
- For **executive summaries, EBC decks, or web copy**, use the whitepaper and TEV landing page language rather than the detailed formula sections.

---

# 2. Core Cisco Retail 2.0 concepts

## 2.1 The central thesis

Retailers do not win the AI-native era by deploying more disconnected tools. They win by connecting intelligence to execution across the store estate.

## 2.2 The problem

The issue is not a lack of technology. The issue is fragmentation: customer data, store systems, associate tools, loss-prevention signals, IT visibility, and AI pilots often operate separately.

## 2.3 The connected store operating model

Cisco Retail 2.0 helps retailers move from fragmented digital transformation to connected store execution through a reusable operating loop:

**Sense → Correlate → Act → Learn → Evolve**

## 2.4 The three priorities

| Priority | Meaning |
|---|---|
| Priority 1: CX Growth and Loyalty | Connect store, app, web, messaging, loyalty, service, product data, and retail media into more continuous customer journeys. |
| Priority 2: EX Workforce, Operations, and Margin Protection | Reduce frontline friction, protect associate capacity, improve service flow, automate repetitive tasks, strengthen store safety, and elevate loss prevention into operational visibility. |
| Priority 3: IT Resilience, Trust, and Protection | Build the secure, observable, AI-ready infrastructure required to scale connected retail without increasing fragility. |

## 2.5 The four TEV layers

| Layer | Name | Treatment |
|---|---|---|
| Layer A | Direct annual operating impact | Measurable annual value from service efficiency, labor capacity, shrink reduction, energy savings, support efficiency, incident reduction, and revenue enablement. |
| Layer B | First-party context engine | Strategic asset-value proxy for enriched profiles, consented messaging identities, product-intent signals, known store sessions, and retail media audience readiness. Not annual revenue, booked cash flow, ROI, or profit. |
| Layer C | Protected value at risk | Downtime, cybersecurity, safety, continuity, and trust exposure that the connected store foundation helps protect. Non-additive by default. |
| Layer D | Future option value | Ability to deploy future AI, edge, computer vision, signage, workflow, automation, and retail media use cases faster and with less rework. Usually qualitative unless roadmap or avoided-rework data exists. |

---


# 3. Quick Reference: Cisco Technology, Use Cases, and Value Drivers

Use this section first when a user asks questions such as: **"what technology enables this use case?"**, **"what value drivers use Webex Connect?"**, **"where does Splunk show up?"**, or **"which Cisco capabilities support margin protection?"**

## 3.1 Technology-to-value-driver crosswalk

| Cisco capability / technology | Primary use cases enabled | Related value drivers | TEV layer relevance |
|---|---|---|---|
| Cisco Catalyst / Meraki wireless | Customer and associate connectivity; app-to-store engagement; guest Wi-Fi; device connectivity; store-mode app experiences | Personalized In-Store Engagement; Secure Connected Store Foundation; Connected Device Reliability | Layer A, Layer B, Layer C, Layer D |
| Cisco Spaces-style location intelligence | Presence, proximity, dwell, traffic, zones, wayfinding, queue visibility, department engagement | Personalized In-Store Engagement; Intelligent Labor and Service Flow; Dynamic Digital Merchandising and Retail Media | Layer A, Layer B |
| Cisco Meraki cameras and sensors | Video visibility, traffic, queue detection, LP review, safety monitoring, operational analytics | Margin Protection and Connected Loss Prevention; Intelligent Labor and Service Flow; Observability; AI-Ready Edge | Layer A, Layer C, Layer D |
| Cisco switching / PoE / PoE++ | Power and connect APs, cameras, sensors, signage, ESL gateways, smart lighting, edge devices | Connected Facilities and Smart Lighting; Secure Connected Store Foundation; Task Automation and ESLs | Layer A, Layer C, Layer D |
| Cisco SD-WAN | Store-to-cloud resilience, SaaS/application connectivity, critical workflow prioritization, store network simplification | Secure Connected Store Foundation; Connected Device Reliability; Observability | Layer A, Layer C |
| Webex Connect | SMS/RCS/WhatsApp/app/web messaging; pickup updates; service recovery; product-originated conversations; scan-to-message; operational alerts | Omnichannel Journey Orchestration; AI-Powered Service and Outreach; Shelf-Edge Scan-to-Learn; AI-Assisted Workflow Protection; Margin Protection workflows | Layer A, Layer B |
| Webex AI | Intent detection, next-best-action, summarization, conversational assistance, routine inquiry automation, associate assist | AI-Powered Service and Outreach; AI-Assisted Workflow Protection; Shelf-Edge Scan-to-Learn; Omnichannel Journey Orchestration | Layer A, Layer B, Layer D |
| Webex Contact Center | Live escalation, agent routing, service continuity, contact-center automation, context-preserving handoffs | AI-Powered Service and Outreach; Omnichannel Journey Orchestration; Shelf-Edge Scan-to-Learn | Layer A, Layer B |
| AppDynamics | Application performance, API monitoring, business transaction visibility, app-to-store journey assurance | Observability as a Business Capability; Connected Device Reliability; Secure Connected Store Foundation; AI-Ready Edge | Layer A, Layer C |
| ThousandEyes | WAN, internet, SaaS, cloud path visibility; outage/path diagnosis; third-party dependency visibility | Observability as a Business Capability; Secure Connected Store Foundation; Connected Device Reliability | Layer A, Layer C |
| Splunk | Telemetry correlation across apps, networks, security, POS, video, operations, identity, incidents | Observability; Margin Protection; Connected Device Reliability; Cyber Resilience; AI-Ready Edge | Layer A, Layer C, Layer D |
| Cisco Duo | MFA, device trust, user access protection, credential risk reduction | Zero Trust and Store Segmentation; Cyber Resilience and Risk Reduction | Layer C |
| Cisco ISE | Identity-based access, device profiling, segmentation, policy enforcement | Zero Trust and Store Segmentation; Secure Connected Store Foundation | Layer C |
| Cisco Secure Access | Secure private app, internet, and cloud access; SASE/SSE posture | Zero Trust and Store Segmentation; Cyber Resilience and Risk Reduction | Layer C |
| Cisco XDR | Detection and response across security telemetry | Cyber Resilience and Risk Reduction | Layer C |
| Cisco Umbrella / DNS-layer security | Block malicious destinations; reduce internet-based exposure; support secure branches | Cyber Resilience and Risk Reduction; Secure Connected Store Foundation | Layer C |
| ESL infrastructure connected through Cisco store infrastructure | Digital price updates, shelf-edge product education, dynamic content, scan-to-learn, execution visibility | Task Automation and ESLs; Dynamic Digital Merchandising and Retail Media; Shelf-Edge Scan-to-Learn | Layer A, Layer B |
| Digital signage / VisionEDGE-style retail media infrastructure | In-store media, dynamic content, supplier-funded activation, inventory-aware signage | Dynamic Digital Merchandising and Retail Media; Personalized In-Store Engagement | Layer A, Layer B |

## 3.2 Value-driver-to-technology crosswalk

| Priority | Value driver | Cisco capabilities that commonly enable it | What to emphasize in answers |
|---|---|---|---|
| CX | Personalized In-Store Engagement | Wireless, Cisco Spaces-style location intelligence, Webex Connect, Webex AI, Webex Contact Center, AppDynamics, ThousandEyes, Splunk | Connect known customer context to physical store presence and action. |
| CX | Omnichannel Journey Orchestration | Webex Connect, Webex AI, Webex Contact Center, AppDynamics, ThousandEyes, SD-WAN, Splunk | Coordinate pickup, returns, messaging, service recovery, escalation, and follow-up. |
| CX | AI-Powered Service and Outreach | Webex Connect, Webex AI, Webex Contact Center, AppDynamics, ThousandEyes, Splunk | Automate routine service while preserving context-rich human escalation. |
| CX | Dynamic Digital Merchandising and Retail Media | Meraki/Catalyst networking, PoE, Cisco Spaces-style location, Meraki cameras/sensors, Webex Connect, AppDynamics, ThousandEyes, Splunk, security | Make in-store content measurable, adaptive, and operationally governed. |
| CX | Shelf-Edge Scan-to-Learn | Webex Connect, Webex AI, Webex Contact Center, store networking, ESL/shelf-edge infrastructure, AppDynamics, ThousandEyes, Splunk, security | Turn product-level intent into a digital service or education flow. |
| EX/Ops | Margin Protection and Connected Loss Prevention | Meraki cameras/sensors, networking/PoE, Splunk, Webex Connect, Webex Contact Center, Cisco security/segmentation, AppDynamics, ThousandEyes | Connect video, POS, inventory, case, and workflow context. Avoid unsafe associate confrontation. |
| EX/Ops | Intelligent Labor and Service Flow | Cisco Spaces-style location, Meraki cameras/sensors, wireless, switching, Webex Connect, Webex AI, Splunk, AppDynamics, ThousandEyes | Detect queues, dwell, pickup friction, and service needs early, then route work. |
| EX/Ops | AI-Assisted Workflow Protection | Webex Connect, Webex AI, Webex Contact Center, Cisco networking, AppDynamics, ThousandEyes, Splunk | Protect associate capacity by deflecting routine, context-dependent interruptions. |
| EX/Ops | Task Automation and ESLs | Meraki/Catalyst networking, PoE, ESL infrastructure, Webex Connect, Splunk, AppDynamics, ThousandEyes, security/segmentation | Reduce manual price/promotion execution and create a shelf-edge engagement foundation. |
| EX/Ops | Connected Facilities and Smart Lighting | Catalyst PoE/PoE++, Meraki networking, sensors/IoT, Splunk, security/segmentation, AppDynamics, ThousandEyes | Connect facilities telemetry to energy, safety, occupancy, and operational visibility. |
| EX/Ops | Connected Device Reliability | Wireless, switching, SD-WAN, Meraki cloud management, AppDynamics, ThousandEyes, Splunk, security | Reduce MTTR, truck rolls, and workflow disruption from device/app/network issues. |
| IT | Secure Connected Store Foundation | Catalyst, Meraki, switching, SD-WAN, PoE, Meraki cloud management, ThousandEyes, Splunk | Store infrastructure is the execution foundation for CX, EX, LP, retail media, facilities, security, and AI. |
| IT | Zero Trust and Store Segmentation | ISE, Duo, Secure Access, segmentation, XDR, Umbrella, Meraki/Catalyst, Splunk | Expand connected-store capabilities without expanding risk at the same pace. |
| IT | Observability as a Business Capability | AppDynamics, ThousandEyes, Splunk, network telemetry, Meraki cloud management, security telemetry | Translate technical performance into customer, associate, and business-workflow impact. |
| IT | Cyber Resilience and Risk Reduction | Duo, Secure Access, ISE, XDR, Umbrella, segmentation, Splunk, Cisco networking | Treat cybersecurity as business continuity and protected value. |
| IT | AI-Ready Edge and Cloud Infrastructure | Networking, switching, SD-WAN, PoE, Meraki cameras/sensors, AppDynamics, ThousandEyes, Splunk, Cisco security, Webex AI/Connect/Contact Center | Create reusable foundation for future AI use cases without rebuilding store architecture. |

## 3.3 Short answer templates for common user questions

### If asked: "What tech enables [use case]?"
Answer with:
1. The primary Cisco capability layer.
2. The enabling technologies.
3. The connected workflow.
4. The value driver and TEV layer.
5. Any caveats about customer inputs, deployment scope, or overlap.

### If asked: "Which use cases does [Cisco product] support?"
Answer with:
1. A short list of applicable value drivers.
2. The operational role of that product.
3. Example workflows.
4. Where value appears: Layer A, B, C, or D.

### If asked to calculate TEV
Use the TEV calculator rules in Section 6. Do not invent customer inputs. Label unknowns. Use conservative defaults. Keep Layer A, B, C, and D separate.

### If asked for whitepaper-style narrative
Use the strategic narrative in Section 4 and the TEV page copy in Section 5. Do not answer like a formula sheet unless the user asks for a scenario or calculation.


---

# 4. Full Whitepaper Narrative

Retail in the AI-Native Era

The Connected Store Operating Model

A Cisco Retail 2.0 Whitepaper

Edited v7.1 draft | May 2026

Retailers do not win the AI-native era by deploying more disconnected tools. They win by connecting intelligence to execution across the store estate.


# Executive Summary


Retailers have not failed to invest in technology. They have modernized ecommerce, launched loyalty programs, deployed mobile apps, upgraded store networks, added self-checkout, expanded fulfillment options, tested AI, rolled out cameras and sensors, invested in retail media, and moved more systems to the cloud.

The problem is not a lack of technology. The problem is fragmentation.

Customer data is fragmented across channels. Store systems are fragmented across devices and workflows. Associate tools are fragmented across tasks. Loss-prevention signals are fragmented across video, POS, inventory, and case management. IT visibility is fragmented across store networks, cloud services, applications, APIs, and third-party providers. AI pilots are often fragmented from the operating systems they need to influence.

AI raises the cost of that fragmentation. AI cannot create trusted outcomes from disconnected systems. It needs reliable data, real-time context, secure identity, resilient connectivity, governed workflows, and operational telemetry. Without those foundations, AI remains a pilot, a dashboard, or a point solution. With them, AI becomes part of execution: sensing what is happening, correlating signals, triggering action, learning from outcomes, and helping the store adapt over time.

This is the shift at the center of Cisco Retail 2.0: the move from fragmented digital transformation to connected store execution.

The modern store is no longer just a place to transact. It is a fulfillment node, service hub, workforce environment, media surface, product discovery layer, data source, security perimeter, and distributed edge environment. It is where customer experience, employee experience, store operations, loss prevention, merchandising, retail media, cybersecurity, and IT resilience converge.

Cisco’s role is to help retailers build the trusted foundation for that operating model. Secure networking, Wi-Fi, switching, SD-WAN, PoE, smart cameras, sensors, electronic shelf label infrastructure, Webex Connect, Webex AI, Webex Contact Center, AppDynamics, ThousandEyes, Splunk, and Cisco security capabilities can help connect the physical and digital business. The value is not in any one technology alone. The value is in connecting the systems that shape retail execution in real time.


|Priority|What it means|
|---|---|
|Priority 1: CX Growth and Loyalty|Connect store, app, web, messaging, loyalty, service, product data, and retail media into more continuous customer journeys.|
|Priority 2: EX Workforce, Operations, and Margin Protection|Reduce frontline friction, protect associate capacity, improve service flow, automate repetitive tasks, strengthen store safety, and elevate loss prevention into operational visibility.|
|Priority 3: IT Resilience, Trust, and Protection|Build the secure, observable, AI-ready infrastructure required to scale connected retail without increasing fragility.|


The economic implications are significant because connected-store capabilities influence multiple value pools at once. A more connected store can improve revenue enablement, service efficiency, associate capacity, shrink reduction, energy usage, support productivity, downtime protection, cybersecurity resilience, retail media monetization, and first-party customer understanding. The value story is strongest when those outcomes are treated as related but distinct: direct operating impact, strategic upside, protected value at risk, and future option value.

This paper uses selective benchmark signals to make the business case concrete without turning the narrative into a calculator. Industry shrink, return fraud, service-cost, retail-media, observability, and cyber-resilience benchmarks help frame why the opportunity matters. The companion TEV landing page should carry the detailed methodology, formulas, source registry, scenario inputs, deployment assumptions, and overlap controls.

The AI-native store is not a hype story. It is an execution story. Retailers that win in this next era will not simply deploy more AI. They will connect the store well enough for AI, automation, analytics, security, and human teams to act together.


# 1. Retail Has Entered the AI Execution Era


Retail’s first wave of AI was defined by experimentation. Retailers tested recommendation engines, demand forecasting, computer vision, workforce analytics, dynamic pricing, fraud models, chatbots, and generative AI assistants. Many efforts produced useful insights; some produced measurable value. But many remained disconnected from the operating core of the business.

A retailer can have an impressive AI pilot and still struggle to execute consistently across stores. A model can recommend the right action, but the value is limited if the data is stale, the workflow is manual, the associate never sees the recommendation, the customer message is not triggered, the application is slow, or the store network cannot support the connected experience.

Retail is now moving from the AI experimentation era to the AI execution era. Execution means AI is embedded into the daily flow of retail work: routing messages, detecting exceptions, adjusting content, recommending next actions, supporting associates, summarizing service interactions, improving investigations, prioritizing tasks, monitoring systems, and reducing friction.

This is why AI-native retail is not simply about AI models. It is about the operating environment that allows AI to act responsibly. AI execution requires trusted data, real-time context, secure identity, reliable connectivity, governed workflows, and systems that can coordinate action.


## The retail technology paradox


Most enterprise retailers have more technology than ever, but many still experience more friction than they should. More devices can mean more support tickets. More customer channels can mean more inconsistent journeys. More data can mean more silos. More dashboards can mean less shared context. More AI pilots can mean more governance risk.

The issue is not whether a retailer has technology. The issue is whether the technology is connected enough to produce repeatable operating value.

In the AI execution era, isolated technology creates diminishing returns. Connected technology compounds.


## From digitization to orchestration


Most retailers have already digitized many parts of the business. The work ahead is orchestration.

Digitization converts analog processes into digital ones. Orchestration connects digital processes so they operate together. A digitized store may have digital signs, mobile apps, ESLs, cameras, loyalty tools, and messaging. An orchestrated store connects those systems into customer journeys, associate workflows, operational alerts, business rules, and measured outcomes.

- A digital sign displays content. An orchestrated sign adapts based on local inventory, daypart, traffic, promotion strategy, and campaign measurement.

- A chatbot answers generic questions. An orchestrated AI assistant begins with product, store, loyalty, inventory, and service context.

- A camera records video. An orchestrated camera estate supports loss prevention, queue detection, safety response, layout analytics, and investigation workflows.

- A network keeps devices online. An orchestrated network provides visibility into which customer, associate, or operational workflows are affected when connectivity degrades.


## Why the buying case changes


In the past, a retailer could evaluate store technologies one at a time: Wi-Fi modernization as infrastructure, digital signage as merchandising, cameras as security, contact center as service, ESLs as operations, and observability as IT. Each project had its own owner, budget, and ROI logic.

That model is becoming less useful because the value pools increasingly overlap. The same camera network can support loss prevention, queue visibility, safety, and operational intelligence. The same messaging platform can support pickup updates, service recovery, product-originated conversations, loyalty engagement, and contact deflection. The same observability layer can protect customer experience, associate productivity, and IT response.

Retailers need a buying case that reflects the connected nature of the work while still separating direct annual impact, strategic upside, protected value, and future option value.


# 2. The Store Is Not Going Away - Its Mission Is Changing


The physical store remains one of retail’s most important assets. It is where customers discover products, confirm choices, pick up orders, return items, compare alternatives, seek help, interact with associates, experience brands, and make immediate purchase decisions.

But the store’s mission has changed.

For decades, the store was primarily designed to present inventory, support service, and complete transactions. That role still matters, but it is no longer enough. Today’s store must also operate as a fulfillment node, service hub, workforce environment, media surface, data source, security perimeter, and distributed technology edge.

That expanded mission creates new pressure. A store now has to support physical shopping and digital fulfillment at the same time. It has to serve customers who may arrive with online research, app-based offers, loyalty status, pickup orders, return authorizations, product questions, and service histories already in motion. It has to give associates enough context to respond quickly. It has to protect margin in an environment where shrink, fraud, and safety are operational concerns. It has to keep devices, applications, cameras, sensors, signage, and networks working continuously. And it has to do all of this without making the experience feel complicated to the customer.

The store is not going away. It is becoming more demanding.


## The store as a fulfillment node


Omnichannel retail has turned stores into active fulfillment infrastructure. Stores now support buy online, pick up in store; curbside pickup; ship-from-store; returns; exchanges; substitutions; local delivery; and inventory-aware customer journeys.

That shift changes the operating model. Store teams are no longer only serving shoppers who are physically present. They are also serving digital demand. A store associate may be helping a customer in aisle three while another customer is waiting in the parking lot for pickup, a third is asking about substitution options through messaging, and a fourth is returning an item purchased online.

For the customer, these are not separate channels. They are one relationship with the retailer. For the store, they are separate workflows unless the underlying systems are connected.

A connected store can make fulfillment feel coordinated. Pickup readiness can trigger a message. Customer arrival can trigger associate tasking. Substitution questions can route through a messaging flow. Inventory exceptions can be escalated before the customer arrives. Returns can be connected to order history. Store teams can see what needs action, not just what has already gone wrong.


## The store as a service hub


As digital channels handle more routine transactions, the store increasingly becomes the place where customers seek confidence, assurance, and resolution.

A shopper may come to the store because they want to compare products physically. They may need help choosing a size, checking compatibility, understanding ingredients, resolving a return, asking about a warranty, or completing a high-consideration purchase. They may have already searched online and still need human reassurance.

This creates an opportunity and a burden. The store can strengthen loyalty when it resolves complexity well. But it can weaken loyalty when customers encounter inconsistent information, unavailable associates, disconnected service histories, or repetitive handoffs.

AI and connected workflows can help, but only if they are grounded in the right context. A product assistant is useful when it knows the exact item, local inventory, customer question, and escalation path. A contact center is more effective when it can see the store context. An associate is more helpful when they have access to product, order, loyalty, and service information in one place.


## The store as a service hub depends on connected context.



## The store as a workforce environment


The store is also where retail work happens.

Associates are being asked to support more channels, more devices, more fulfillment complexity, more safety concerns, more product questions, and more exceptions. They are expected to deliver human service while also executing tasks generated by digital systems.

When the operating model is fragmented, associates absorb the friction. They search across systems, repeat information, troubleshoot devices, manage customer frustration, walk the store to check conditions, manually update labels, handle pickup exceptions, and work around disconnected tools.

That friction is not just an employee experience issue. It is a business performance issue. Every minute spent navigating broken workflows is a minute not spent serving customers, fulfilling orders, preventing loss, or improving store execution.

The connected store should make work easier to execute. It should give associates better context, reduce repetitive questions, route exceptions clearly, automate routine tasks, and keep critical devices reliable.


## The store as a media and merchandising surface


Retail media and digital merchandising are also changing the store’s mission.

The store is no longer only a place where merchandising is physically set and periodically changed. It is becoming a dynamic content environment. Digital signage, electronic shelf labels, mobile prompts, in-store audio, shelf-edge codes, and app experiences can help retailers adapt content by store, department, inventory, time of day, customer mission, supplier campaign, or local event.

That creates new monetization and engagement opportunities. But it also creates a risk: the store can become cluttered, irrelevant, or confusing if digital content is not connected to customer needs and local operating reality.

The best in-store media does not simply add more screens. It makes the store more useful. It can guide customers, highlight relevant offers, explain products, promote available inventory, support supplier-funded education, and connect physical engagement to measurable outcomes.


## The store as a data source


Ecommerce has long had a data advantage. Retailers can see searches, clicks, carts, abandonments, conversions, and engagement paths. Physical retail has historically been less visible.

Connected stores narrow that gap.

Wi-Fi, location intelligence, cameras, sensors, POS, mobile apps, loyalty engagement, ESLs, digital signage, shelf-edge 2D codes, associate devices, and operational telemetry can help retailers understand what is happening before, during, and after the transaction.

That visibility can answer important questions:

- Where do customers dwell but not convert?

- Which departments need more support?

- Which products generate questions?

- Where do queues build?

- Which promotions are active when inventory is unavailable?

- Which store workflows are creating repeated service issues?

- Which devices or applications are slowing associates down?

- Which shrink patterns repeat across locations?

- Which fulfillment steps create customer friction?

The point is not to collect data for its own sake. The point is to use store signals to improve execution.


## The store as a security perimeter


The connected store also expands the security perimeter.

Every new device, application, vendor connection, customer channel, and data flow creates potential exposure. POS systems, payment devices, cameras, sensors, digital signage, associate handhelds, guest Wi-Fi, smart lighting, HVAC, refrigeration, ESLs, kiosks, and cloud applications all need appropriate access, segmentation, monitoring, and policy enforcement.

Physical security and cybersecurity are also becoming more connected. A compromised device can create operational risk. A disconnected camera network can limit safety response. A weak vendor access policy can expose store systems. A flat network can increase blast radius. A customer data flow without proper governance can create privacy and trust risk.

The store is now both a physical and digital risk environment.


## The store as a distributed edge


Finally, the store has become a distributed edge environment.

It hosts devices, sensors, cameras, applications, local workflows, real-time data, and cloud dependencies. Some decisions need to happen locally. Some need cloud scale. Some require both. AI use cases such as computer vision, queue detection, product assistance, dynamic signage, smart lighting, and operational automation all depend on the ability to connect edge signals to enterprise systems securely and reliably.

This changes how retailers should think about infrastructure. Store infrastructure is not merely a cost center. It is the foundation that allows the store to fulfill its new mission.

A retailer cannot deliver AI-native customer experience, associate productivity, loss prevention, retail media, and IT resilience on disconnected or fragile infrastructure.

The modern store has become too important, too complex, and too connected to be managed as a set of isolated systems. It needs an operating model built for the role it now plays.


# 3. The Connected Store Operating Model


The connected store operating model is the bridge between retail’s AI ambition and retail’s daily execution. It is not a single product, platform, application, or dashboard. It is a way of organizing the store so that customer signals, associate workflows, operational events, technology telemetry, and business systems can work together in real time.

For many retailers, the challenge is that digital tools operate in parallel. The app knows one version of the customer. The loyalty platform knows another. The store team sees only part of the journey. Cameras may capture video but not connect to POS events, safety workflows, or operational analytics. IT tools may monitor systems but not show which customer or associate journeys are affected when performance degrades.

The connected store turns those isolated capabilities into a coordinated system with five jobs: sense what is happening, correlate signals into useful context, act through the right workflow, learn from outcomes, and evolve as conditions change.


## Retailers enter the model at different points


Not every retailer starts from the same place. Some are integrating capabilities they have already deployed; for them, the work is orchestration. Others are establishing the foundation for the first time; for them, the work begins with reliable, secure, observable infrastructure on which connected experiences can be built.

Both groups need the same end-state: a store environment where connectivity, observability, security, engagement, and operational workflows compound over time. But the path is not a single-step transformation. Retailers that try to install AI-native experiences on top of unstable connectivity, ungoverned devices, unmanaged cameras, or flat networks often find that the experience layer cannot perform reliably enough to produce business value.


## From systems of record to systems of action


Retailers have spent years building systems of record: customer records, transaction records, product records, inventory records, employee records, and incident records. Those systems remain essential. But AI-native retail requires systems that help the business act while the moment is still unfolding.

A pickup delay should not become visible only after the customer complains. A queue should not be addressed only after the line is obvious. A self-checkout exception should not require an unsafe or inconsistent associate response. A digital sign should not keep promoting unavailable inventory. A customer product question should not restart as a generic chatbot interaction when the exact SKU, store, and aisle are already known.


## Connection must be technical, operational, and measurable


- Technical connection: devices, applications, networks, APIs, cloud services, and data systems must communicate reliably and securely.

- Operational connection: the business must define what happens when a signal appears, who receives the task, and when AI should act or escalate.

- Measurement connection: outcomes must be captured so the retailer knows whether the action reduced friction, saved time, improved conversion, reduced loss, or protected uptime.

The purpose of the connected store is not to make the store feel more technical. It is to make it feel more coordinated: easier for customers, clearer for associates, more visible for store leaders, more manageable for IT, and more valuable for the business.


# 4. SCALE: Sense, Correlate, Act, Learn, Evolve


SCALE provides a simple operating language for the connected store: Sense, Correlate, Act, Learn, and Evolve. It explains how a store moves from fragmented signals to coordinated execution.

A retailer does not become AI-native because it has an AI model. It becomes AI-native when intelligence is connected to action.


|SCALE stage|What it means in the store|
|---|---|
|Sense|Capture useful signals from customers, associates, cameras, sensors, POS, inventory, ESLs, signage, mobile apps, loyalty, CPaaS, contact center, applications, APIs, network telemetry, security events, and facilities systems.|
|Correlate|Turn signals into context by connecting events across staffing, traffic, POS, video, inventory, applications, devices, identity, security, and store workflows.|
|Act|Route context into the right workflow: customer message, associate task, digital signage change, contact center escalation, LP review, IT remediation, security policy action, or facilities adjustment.|
|Learn|Measure whether the action reduced wait time, improved pickup completion, lowered contact volume, reduced investigation time, improved engagement, saved energy, or shortened resolution time.|
|Evolve|Use the same foundation to support new use cases as customer expectations, store formats, AI capabilities, retail media models, security threats, and device ecosystems change.|


A digitized store has connected devices. An AI-native connected store has a loop. That loop is what allows customer experience, employee experience, loss prevention, merchandising, facilities, security, and IT operations to work from the same foundation.


# 5. Why Cisco


Most retail friction is not created by the absence of technology. It is created by fragmentation between technologies. That is why Cisco has a distinctive role in the AI-native retail conversation.

Retailers do not need another disconnected point solution. They need the systems already shaping the store experience to work together: networks, devices, applications, cameras, sensors, messaging, contact center, security, observability, cloud services, associate tools, and operational workflows.

Cisco’s role is not to replace the retailer’s commerce platform, loyalty system, POS, order management system, product information system, workforce platform, or retail media stack. Cisco’s role is to help connect, secure, observe, and orchestrate the environment in which those systems operate.

The stronger Cisco argument is not: Cisco provides store networking. The stronger argument is: Cisco helps retailers move from fragmented technology deployments to connected execution.


## Cisco as the connective tissue of the AI-native store


Store devices need reliable connectivity. Applications need performance assurance. Customer journeys need messaging orchestration. Associates need context and tasking. Cameras need network, power, policy, storage, analytics, and workflow integration. Electronic shelf labels, signage, and shelf-edge systems need connectivity and management. Security requires identity, segmentation, and telemetry. IT teams need visibility across local stores, SaaS, cloud, WAN, APIs, and third-party services. AI needs governed access to real-time context.

Cisco is structurally positioned across these domains. The network remains foundational, but the retail opportunity is broader: secure infrastructure, observability, customer engagement orchestration, contact center, AI-enabled workflows, cameras, sensors, identity, segmentation, and telemetry.


## Why this matters to retail executives



|Executive|Cisco relevance|
|---|---|
|CIO|Reduce complexity across distributed stores, cloud services, applications, devices, identities, and security domains.|
|COO|Make store execution more visible, connected, and responsive.|
|CFO|Connect technology investment to value pools such as labor productivity, shrink reduction, downtime avoidance, risk reduction, energy savings, contact deflection, and revenue enablement.|
|Chief Digital Officer|Connect digital intent to physical execution through messaging, app performance, location context, service orchestration, and store infrastructure.|
|Chief Store Officer|Reduce frontline friction by improving device reliability, task routing, service flow, safety visibility, and operational awareness.|
|CISO|Protect the expanding connected-store attack surface through identity, segmentation, secure access, threat detection, telemetry, and policy enforcement.|
|Head of CX|Make customer journeys feel more continuous across store, app, messaging, pickup, service, loyalty, and contact center.|



## Three priorities, one connected foundation


Cisco Retail 2.0 should organize the business case around three priorities: CX Growth and Loyalty; EX Workforce, Operations, and Margin Protection; and IT Resilience, Trust, and Protection.

These priorities should not be treated as separate technology tracks. They are three views of the same connected store. The same foundation that improves customer engagement can also improve associate workflows. The same camera estate that supports loss prevention can also support queue visibility and safety. The same observability layer that reduces IT resolution time can also protect digital customer journeys. The same secure network that supports POS can also support sensors, signage, ESLs, smart lighting, and AI workloads.


# 6. Priority 1 - CX Growth and Loyalty


Retail growth in the AI execution era will increasingly depend on how well retailers connect digital intelligence to physical moments.

Customers no longer experience store, app, service, loyalty, and fulfillment as separate channels. They experience them as one relationship with the brand. When that relationship feels fragmented - irrelevant offers, inconsistent communications, poor in-store guidance, unavailable product information, disconnected pickup updates, repetitive service interactions, or handoffs that lose context - loyalty weakens quickly. When it feels connected, customers make decisions more confidently, complete journeys more easily, and return more often.

This matters because stores remain central to retail economics even as customer expectations are increasingly shaped by digital convenience. Shoppers want the speed, relevance, and transparency of ecommerce with the immediacy, assurance, and human support of physical retail. They expect promotions to be timely, inventory to be accurate, product information to be accessible, service to be proactive, and transitions between web, mobile, messaging, contact center, and store to feel seamless.

For retailers, that creates a clear strategic opportunity: use data, location intelligence, AI, messaging, product information, and connected workflows to make the store journey more relevant, more helpful, and easier to complete. The goal is not to make every moment feel high-tech. It is to make every moment feel more intuitive.

Cisco’s role in this priority is to help connect the systems that shape customer experience in real time: secure wireless connectivity, location awareness, store analytics, messaging orchestration, application assurance, digital signage infrastructure, AI-enabled service flows, contact center escalation, and omnichannel coordination. Together, these capabilities can help retailers reduce friction, improve engagement, and turn disconnected touchpoints into a more continuous customer relationship.

Within this priority, five value drivers matter most:


|Value driver|What it does|
|---|---|
|Personalized In-Store Engagement|Connects store presence, loyalty, app, Wi-Fi, and customer context to make the physical journey more relevant|
|Omnichannel Journey Orchestration|Coordinates store, app, pickup, service, messaging, loyalty, and contact center interactions|
|AI-Powered Service and Outreach|Automates routine service moments and escalates complex needs with context|
|Dynamic Digital Merchandising and Retail Media|Uses signage, ESLs, media surfaces, inventory, traffic, and content to make the store more adaptive and measurable|
|Shelf-Edge Scan-to-Learn|Optional connected CX use case that uses SKU-level 2D codes to launch product education, offers, AI Q&A, or service flows|



## Value Driver 1: Personalized In-Store Engagement



### What challenge it solves


Many retailers know their customers well in digital channels but lose that intelligence when the customer enters the store.

A shopper may be recognized in the app, targeted through email, enrolled in loyalty, and actively browsing products online. But once that same shopper walks into a physical location, the experience can become generic. The store may not know what the customer has researched, which offers are relevant, what products are available locally, or where the shopper is likely to need help. The customer experiences the brand as disconnected: personalized online, anonymous in person.

That gap matters because physical shopping often involves moments where relevance and confidence influence the sale. A shopper may need to compare products, find an item quickly, validate a promotion, check inventory, understand a size or specification, or decide between a national brand and a private-label alternative. When the retailer can connect digital intent to physical context, the store becomes more useful. When it cannot, the customer is forced to do the work.

Personalized in-store engagement is not about making the store feel intrusive. It is about making the physical journey feel more aware, timely, and helpful.


### Use cases Cisco enables


- Arrival-aware engagement: When an opted-in customer or app user enters a store, the retailer can shift the experience into store mode with relevant offers, pickup instructions, or shopping-list context.

- Location-based offers: Store zone, department, or dwell context can help the retailer present offers that match the customer’s current mission rather than generic promotions.

- App-to-store continuity: Digital research, wishlists, loyalty preferences, and shopping lists can carry into the in-store experience so customers do not restart their journey.

- In-store wayfinding: Customers can find products, departments, pickup areas, service counters, or promotional zones more easily.

- Dwell and traffic tuning: Retailers can understand where customers spend time and where departments may need better content, staffing, or merchandising.

- Known-customer recapture: If a known shopper engages but does not purchase, the retailer can follow up later through permissioned channels with relevant content or offers.

- Personalized service triggers: High-value or high-friction moments can be routed to the right associate or support channel with customer and store context.


### Technologies used


Cisco Spaces-style location intelligence helps retailers understand store presence, movement, dwell, and location-based context. It can support store-mode experiences, traffic analysis, wayfinding, and zone-level engagement.

Cisco Meraki and Catalyst wireless provide the connectivity layer required for guest, associate, and operational experiences. Reliable Wi-Fi is foundational for app engagement, associate devices, store analytics, and connected customer journeys.

Webex Connect can orchestrate opt-in customer messaging across SMS, RCS, WhatsApp, app chat, web chat, and other digital channels. It helps retailers turn context into timely engagement.

Webex AI can support intent classification, next-best-action logic, conversational assistance, and service automation.

Webex Contact Center can escalate higher-complexity needs with customer, store, and interaction context intact.

AppDynamics and ThousandEyes help assure the digital experiences that support app-to-store journeys, including application performance, API dependencies, and internet path visibility.

Splunk can correlate store, customer, application, network, and operational telemetry to help retailers understand what is happening across the journey.


### Business impacts


Personalized in-store engagement creates measurable value across conversion, relevance, loyalty, and first-party customer understanding.

- Up to 40% higher in-store conversion when shoppers use digital tools during the store journey.

- Vendor-commissioned customer-data TEI studies have also reported conversion and messaging gains when customer profiles are unified and activated across channels.

- Greater relevance at the moment of decision by connecting app, loyalty, Wi-Fi, location, and store context.

- Stronger continuity between digital intent and physical behavior, including store-mode app experiences, wayfinding, offers, and post-visit follow-up.

- More useful first-party context from physical shopping journeys, including known sessions, product interest, and consented engagement signals.

At an executive level, this value driver matters because it helps the retailer turn store traffic into measurable customer context. Instead of treating the physical visit as a blind spot, the business gains a way to connect customer presence with digital engagement, operational insight, and growth outcomes.


### Illustrative example


A sporting goods customer researches running shoes online, saves two models to a wishlist, and later enters a store. The app shifts into store mode, showing aisle location, available sizes, and a loyalty offer on one of the models. Store traffic data indicates the footwear department is busy, so a floor associate receives a lightweight task prompt to check the area. The customer scans a shelf label to compare cushioning and size availability, then receives a Webex-powered prompt offering help. The experience feels personal without requiring the shopper to restart the journey.


## Value Driver 2: Omnichannel Journey Orchestration



### What challenge it solves


Omnichannel friction occurs when the customer’s journey moves faster than the retailer’s systems.

A customer buys online but pickup status is unclear. A substitution requires approval but the message never reaches the customer. A return begins in the app but has to be re-explained at the service counter. A loyalty offer appears in email but fails at checkout. A contact center agent resolves an issue but the store cannot see the context. Each breakdown may look like a small operational issue, but to the customer it feels like the retailer does not know them.

This is especially damaging because omnichannel customers often represent some of the most valuable relationships in retail. They engage across more touchpoints, generate more data, and expect the retailer to coordinate experiences across channels. When that coordination fails, the retailer risks losing not just one transaction, but trust in the broader relationship.

Omnichannel journey orchestration is the discipline of making those transitions feel connected.


### Use cases Cisco enables


- Pickup and curbside coordination: Customers can receive readiness alerts, arrival instructions, substitution prompts, pickup updates, and service recovery messages.

- Returns orchestration: Return eligibility, instructions, receipts, and service counter context can move across app, messaging, and store workflows.

- Service recovery: Fulfillment delays, failed offers, out-of-stock events, or pickup issues can trigger proactive communication before the customer complains.

- Cross-channel escalation: A customer can move from messaging to contact center or store support without losing context.

- Order-status automation: Routine order, delivery, pickup, and return questions can be deflected through AI-assisted messaging.

- Post-visit follow-up: Service, loyalty, survey, replenishment, or support messages can be triggered based on completed journeys.

- Associate-aware handoff: Store teams can receive customer, order, product, and issue context when a digital journey requires physical action.


### Technologies used


Webex Connect orchestrates customer communications across channels and supports proactive messaging workflows for pickup, returns, substitutions, service recovery, and follow-up.

Webex AI helps classify intent, summarize interactions, automate routine journey steps, and support next-best-action logic.

Webex Contact Center provides live-agent escalation and preserves context across digital-to-human handoffs.

AppDynamics monitors customer-facing applications and APIs involved in order, pickup, loyalty, and service journeys.

ThousandEyes provides visibility into internet, SaaS, and network paths that can affect customer-facing digital experiences.

Cisco networking and SD-WAN support reliable store-to-cloud and store-to-enterprise connectivity.

Splunk can correlate operational events across orders, applications, messages, systems, and incidents.


### Business impacts


Omnichannel journey orchestration creates value by reducing the friction customers feel when store, app, pickup, returns, service, and contact-center workflows do not move together.

- A vendor-commissioned Forrester TEI study of unified customer data activation reported a 15% mobile conversion-rate lift, reinforcing the value of connected customer context across digital and physical journeys.

- Fewer avoidable service contacts from proactive pickup, substitution, order-status, return-status, and service-recovery messaging.

- Improved pickup and curbside experiences by connecting readiness, arrival, substitution, associate tasking, and customer communication.

- Reduced failed handoffs when customers move from app to store, messaging to contact center, or digital service to physical action.

- Higher journey completion and customer confidence when the retailer preserves context across web, mobile, messaging, store, and service channels.

At an executive level, this value driver matters because omnichannel customers do not judge the retailer by channel boundaries. They judge whether the experience works as one relationship. Cisco helps make that relationship more continuous by connecting messaging, service, application assurance, store connectivity, and operational context.


### Illustrative example


A beauty customer orders a product for curbside pickup. Before arrival, one item becomes unavailable. Instead of discovering the issue at the store, the customer receives a Webex Connect message with approved substitutions. The customer selects an alternative, the order updates, and the associate handheld receives the revised pick task. If the customer asks a product question, Webex AI handles the first response and escalates to Webex Contact Center if needed. The customer experiences one coordinated retailer, not a set of disconnected systems.


## Value Driver 3: AI-Powered Service and Outreach



### What challenge it solves


Retail service teams and store associates spend significant time answering routine questions. Customers ask about order status, product availability, store hours, return policies, pickup timing, loyalty offers, warranties, product details, and service options. Many of these questions are repetitive, but they still consume human capacity because the answers depend on context.

Generic automation often fails because it lacks that context. A chatbot that does not know the customer’s order, store, product, loyalty status, or current issue can create more frustration than value. A message that is not connected to fulfillment status can confuse the customer. A contact center handoff that loses conversation history forces the customer to repeat themselves.

AI-powered service and outreach is most valuable when it is connected to real retail context and has clear escalation paths.


### Use cases Cisco enables


- Routine inquiry automation: Common questions can be answered through Webex-powered messaging or AI-assisted flows.

- Product-originated conversations: A customer can begin a support interaction from a product, shelf-edge code, app experience, or digital sign.

- Context-rich escalation: Complex issues can move to a live agent with order, product, store, and conversation context intact.

- Proactive service alerts: Customers can receive timely updates about pickup delays, substitutions, returns, appointments, or service changes.

- Associate-assist workflows: Store associates can receive AI-supported answers for product, policy, order, or inventory questions.

- Post-interaction summaries: AI-generated summaries can reduce manual note-taking and improve continuity across channels.

- Customer preference capture: Service moments can update permissions, preferences, and loyalty context when consent is appropriate.


### Technologies used


Webex Connect routes service interactions across SMS, RCS, WhatsApp, app chat, web chat, and other channels. It supports proactive outreach and two-way service flows.

Webex AI supports conversational assistance, intent detection, summarization, next-best-action logic, and automation of routine steps.

Webex Contact Center handles human escalation, agent workflows, and service continuity when automation is not enough.

Cisco networking provides reliable in-store and cloud connectivity for service interactions.

AppDynamics and ThousandEyes help assure application and API performance during customer service journeys.

Splunk supports event correlation, service analytics, and operational visibility across service, messaging, application, and infrastructure data.


### Business impacts


AI-powered service and outreach creates value by shifting routine, context-dependent interactions away from manual handling while preserving human escalation for higher-value moments.

- Directional service benchmarks show a wide cost gap between self-service interactions and assisted support, with planning benchmarks often using roughly $1.84 for self-service versus $13.50 for assisted support.

- Conversational-commerce TEI evidence has reported a 550% engagement lift and conversion-rate improvement from 3.3% to 12.2% in a WhatsApp-led commerce scenario.

- Lower routine contact volume for order status, pickup, return policy, store hours, loyalty, and product-availability questions.

- Faster response and improved consistency when AI has access to order, product, store, loyalty, and service context.

- Improved escalation quality because complex issues move to agents with conversation, product, customer, and store context intact.

At an executive level, this value driver matters because service automation is only valuable when it reduces friction instead of adding another disconnected channel. Cisco helps connect AI, messaging, contact center, application performance, and store context so automation can improve both cost-to-serve and customer experience.


### Illustrative example


A home improvement shopper scans a shelf tag for a power tool and asks whether it works with a specific battery system. Webex AI answers using governed compatibility data. The shopper then asks about warranty coverage, which depends on purchase history and service terms, so the interaction escalates to Webex Contact Center. The live agent receives the product, store, scan, and conversation context. What could have been a generic chatbot exchange becomes a product-specific service moment.


## Value Driver 4: Dynamic Digital Merchandising and Retail Media



### What challenge it solves


Physical stores have historically been harder to adapt and measure than digital channels. Promotions may not reflect local inventory. Signs can remain static even when traffic patterns change. Supplier-funded media often lives in ecommerce channels while the store, where many decisions still happen, remains under-measured. Retailers may know what was purchased, but not always what was considered, compared, scanned, or ignored.

At the same time, the store can become cluttered if digital content is added without context. More screens do not automatically create better engagement. More promotions do not automatically create better customer experience. Dynamic digital merchandising only works when the content is relevant, timely, measurable, and connected to the operating reality of the store.

The opportunity is to make the store more adaptive without making it feel noisy.


### Use cases Cisco enables


- Inventory-aware signage: Digital content can change when promoted products are low, out of stock, or substituted.

- Daypart and mission-based messaging: Content can adapt for morning, evening, weekend, seasonal, event-driven, or mission-specific traffic.

- In-store media activation: Supplier-funded content can appear on screens, endcaps, shelf-edge experiences, or mobile prompts.

- Private-label promotion: Relevant alternatives can be promoted where appropriate and compliant.

- Traffic-informed merchandising: Dwell, traffic, and zone analytics can inform placement, content, and campaign strategy.

- Campaign attribution: Exposure, scans, loyalty, and transaction signals can be connected where governance permits.

- Operational content: Signage can support wayfinding, service updates, pickup instructions, safety messages, or store announcements.


### Technologies used


Cisco Meraki and Catalyst networking connect digital signage, ESL infrastructure, cameras, sensors, and store systems.

PoE switching can support connected displays, sensors, access points, cameras, and selected infrastructure deployments where appropriate.

Cisco Spaces-style location intelligence helps retailers understand traffic, dwell, and zone-level engagement.

Meraki cameras and sensors can support traffic, queue, and environmental visibility where appropriate.

Webex Connect enables scan-to-message, offer, service, and follow-up workflows tied to in-store engagement.

AppDynamics, ThousandEyes, and Splunk support assurance and performance visibility for signage, APIs, content systems, and campaign-related workflows.

Cisco security capabilities help protect connected signage, devices, data flows, and store networks.


### Business impacts


Dynamic digital merchandising and retail media create value by making the physical store more adaptive, measurable, and monetizable without making the experience feel cluttered.

- Market proof points show in-store retail media moving from experiment to scale: Grocery TV has reported more than 6,500 stores on its in-store media platform, while Albertsons has moved digital screens into more than one-third of its stores.

- IAB in-store retail media standards are giving retailers a common language for impressions, audiences, measurement, and outcomes.

- More relevant promotions when signage, shelf-edge content, and digital surfaces respond to local inventory, daypart, traffic, and shopper mission.

- Reduced wasted impressions from out-of-stock or poorly timed promotions.

- Stronger campaign measurement when exposure, scan, loyalty, traffic, and transaction signals can be connected appropriately.

At an executive level, this value driver matters because the store can become both a merchandising environment and a measurable media environment. Cisco helps connect the infrastructure, location intelligence, signage, security, Webex engagement, and observability layers that make that opportunity operational.


### Illustrative example


A grocery store promotes a seasonal beverage on entrance screens, but local inventory drops below the campaign threshold. Instead of continuing to promote an unavailable item, the content shifts to an alternative product funded by the same supplier campaign. At the same time, dwell data shows strong traffic near prepared foods, so a dinner-bundle promotion is pushed to nearby signage. Campaign performance can later be reviewed against scan, loyalty, and transaction data.


## Value Driver 5: Shelf-Edge Scan-to-Learn



### What challenge it solves


Many purchase decisions stall because customers lack product-level confidence in the aisle. They may need allergen details, compatibility guidance, recipes, reviews, size help, accessory recommendations, or a coupon. They may want to know whether a product works with another item, whether an ingredient is safe for their household, whether there is a better alternative, or whether help is available.

Supplier-owned GS1 2D packaging adoption will expand over time, but retailers do not need to wait for every package to change. Shelf-edge 2D codes on ESLs, paper labels, endcaps, or digital signage can create an immediate product-level entry point controlled by the retailer.

This is an optional supporting CX use case, not a standalone pillar. Its importance is that it demonstrates how the connected store can turn physical intent into a digital, measurable, AI-assisted journey.


### Use cases Cisco enables


- Product education: Customers can launch product details, reviews, videos, instructions, or usage guidance.

- Allergen or ingredient lookup: Sensitive questions can retrieve governed product data rather than rely on generic AI output.

- Compatibility checking: Customers can compare electronics, accessories, tools, parts, or product bundles.

- Recipe and solution building: Grocery, home, beauty, and specialty retailers can recommend complementary items or shopping lists.

- Offer clipping: Shoppers can connect to loyalty offers, rebates, or supplier-funded coupons.

- Associate assistance: Customers can request help with product, aisle, and store context attached.

- Product-originated conversation: A Webex-powered interaction can begin from a specific SKU and physical location.


### Technologies used


Webex Connect routes post-scan interactions across SMS, RCS, WhatsApp, app chat, web chat, and other channels.

Webex AI supports product-specific assistance, intent classification, guided next steps, and conversational automation.

Webex Contact Center escalates complex or sensitive questions to live support with product and store context.

Cisco networking provides secure store connectivity for scan, app, and messaging experiences.

ESL and shelf-edge infrastructure provide the physical entry point for SKU-level digital experiences.

AppDynamics, ThousandEyes, and Splunk help assure and observe the post-scan digital journey.

Cisco security capabilities protect identity, data, devices, and channel access.


### Business impacts


Shelf-edge scan-to-learn creates value by turning product-level questions in the aisle into measurable, AI-assisted engagement.

- GS1 Digital Link and the industry move toward 2D barcodes by 2027 show that product-level digital engagement is becoming a mainstream retail capability, not a niche experiment.

- More confident purchase decisions when shoppers can access product details, allergen information, compatibility guidance, recipes, reviews, offers, or associate help from the shelf edge.

- Reduced repetitive associate questions in categories where product information, availability, fit, compatibility, or policy guidance is frequently needed.

- New product-intent signals that show which SKUs, categories, and departments generate questions before purchase.

- Optional supplier-funded education and retail media opportunities when shelf-edge engagement is governed and measurable.

At an executive level, this value driver matters because it gives retailers a retailer-controlled way to make the aisle more useful before every package carries its own connected experience. The value depends on scan volume, product-data readiness, opt-in, and consent rules, but the strategic opportunity is clear: product interest becomes a measurable interaction instead of an invisible moment.


### Illustrative example


An electronics shopper scans a shelf label for a camera and asks whether a specific memory card is compatible. The experience begins with the exact SKU and store context. Webex AI checks governed compatibility data and provides an answer. The shopper then asks about warranty coverage, which escalates to Webex Contact Center. The retailer captures product-level intent while helping the customer make a confident decision.


# 7. Priority 2 - EX Workforce, Operations, and Margin Protection


Customer experience depends on employee experience. Store teams are the execution layer of the retail promise. They are the people who help customers make decisions, fulfill digital orders, resolve exceptions, manage returns, restock shelves, monitor self-checkout, respond to safety concerns, and keep the store operating when systems do not work as expected.

In many retail environments, associates are also the hidden integration layer. When systems are fragmented, associates compensate. They search for product information across multiple tools. They walk the floor to verify inventory. They answer repetitive customer questions. They manually update price labels. They troubleshoot devices. They re-explain issues that should have carried over from the app or contact center. They respond to queues only after customers are already frustrated. They manage loss-prevention and safety concerns with incomplete context.

That is not sustainable.

The AI execution era should make store work easier to execute, not more complicated. The goal is not simply to add more associate tools. It is to reduce avoidable friction in the work itself: fewer disconnected tasks, fewer manual checks, fewer repetitive questions, fewer unclear escalations, fewer device failures, and fewer situations where associates are asked to solve problems that technology should have prevented or routed earlier.

This priority also includes margin protection. Shrink, fraud, and safety are no longer isolated loss-prevention issues. They are operating-model issues. A store cannot deliver strong employee experience if associates feel unsafe, if self-checkout exceptions are handled inconsistently, if inventory accuracy is poor, or if theft and fraud erode margin without connected visibility. In a connected store, loss prevention becomes part of operational intelligence.

Cisco’s role in this priority is to help connect the systems that shape store execution: wireless connectivity, associate devices, cameras, sensors, POS context, Webex workflows, task routing, observability, security, PoE infrastructure, smart lighting, electronic shelf labels, and analytics. Together, these capabilities can help retailers improve workforce productivity, protect associate capacity, strengthen safety, reduce shrink, and make store operations more visible.

Within this priority, six value drivers matter most:


|Value driver|What it does|
|---|---|
|Margin Protection and Connected Loss Prevention|Connects cameras, POS events, safety workflows, and analytics to reduce shrink and improve operational visibility|
|Intelligent Labor and Service Flow|Uses traffic, queue, task, and fulfillment signals to improve store execution and service response|
|AI-Assisted Workflow Protection|Deflects routine questions and protects associate capacity|
|Task Automation and ESLs|Reduces manual price, promotion, and shelf-execution work|
|Connected Facilities and Smart Lighting|Uses connected lighting, sensors, PoE, occupancy, and facilities telemetry to improve energy, safety, and operations|
|Connected Device Reliability|Improves uptime and performance for handhelds, POS, scanners, cameras, signage, and associate tools|



## Value Driver 1: Margin Protection and Connected Loss Prevention



### What challenge it solves


Shrink, fraud, and safety have become executive-level retail issues because they affect more than the loss-prevention budget. They affect margin, inventory accuracy, associate experience, customer trust, store execution, and brand reputation.

Traditional loss prevention often operates after the fact. A suspicious transaction occurs, video is reviewed later, evidence is assembled manually, and patterns may only become visible after repeated losses. That reactive model is no longer enough. Retailers are dealing with external theft, organized retail crime, internal theft, sweethearting, self-checkout missed scans, barcode switching, return fraud, claims abuse, process failures, and safety incidents. Many of these events cut across POS systems, video systems, inventory systems, staffing patterns, and store workflows.

A connected store gives retailers a better model. Loss-prevention signals can become part of real-time operational visibility. Cameras can support both security and service flow. POS events can be correlated with video. Safety workflows can route alerts without encouraging unsafe associate confrontation. Investigation teams can search, review, and act with more context. Store leaders can understand where shrink, process breakdowns, and customer friction intersect.

Connected loss prevention should not be framed only as catching theft. It is about protecting margin, improving visibility, and making stores safer and easier to operate.


### Use cases Cisco enables


- POS/video correlation: Link transaction exceptions to video events so investigation teams can review context more quickly and consistently.

- Self-checkout anomaly review: Identify missed scans, barcode switching, walkaways, or repeated exception patterns without relying only on manual observation.

- Internal theft and sweethearting detection: Connect exception reporting, transaction data, and video context to investigate suspicious employee-assisted transactions.

- Return fraud and claims review: Correlate returns, receipts, customer history, video, and case data where policy and governance permit.

- Safety and de-escalation workflows: Route alerts to appropriate teams without pushing frontline associates into unsafe confrontation.

- Investigation acceleration: Reduce time spent searching video, assembling evidence, and reconstructing events.

- Dual-use camera analytics: Use the camera estate for queue visibility, safety monitoring, layout intelligence, and operational review in addition to security.

- Inventory accuracy support: Connect shrink signals to phantom inventory, out-of-stock conditions, and fulfillment failures.

- Cross-store pattern detection: Identify repeated events, high-risk zones, or recurring process failures across locations.


### Technologies used


Meraki cameras and sensors provide cloud-managed video, event capture, and store visibility. They can support security, safety, queue visibility, and operational review use cases.

Cisco networking and PoE connect and power camera, sensor, and security infrastructure across distributed store environments.

Splunk can correlate POS, video, security, inventory, and operational events so teams can identify patterns and investigate faster.

Webex Connect can support alerting, workflow routing, and communications across store teams, LP teams, and operational support functions.

Webex Contact Center can support remote escalation, investigation support, customer communication, or associate support workflows where appropriate.

Cisco security and segmentation help protect camera networks, POS-related systems, operational devices, and sensitive investigation workflows.

AppDynamics and ThousandEyes help assure applications, APIs, and network paths connected to LP, POS, and store operations.


### Business impacts


Connected loss prevention creates measurable value across shrink, fraud, investigation productivity, safety response, inventory accuracy, and operational visibility.

- NRF/LPRC reported average retail shrink of 1.6% of sales, representing $112.1B in U.S. retail shrink in the 2023 National Retail Security Survey.

- Appriss Retail / Deloitte research has framed return fraud and abuse as a $100B+ challenge, with return fraud and abuse representing a material share of total return value.

- ECR Retail Loss Group research supports treating self-checkout as a distinct loss-exposure area when fixed SCO, Scan & Go, or BYOD checkout is in scope.

- Faster investigations when POS exceptions, video, inventory signals, case data, and store context can be correlated.

- Broader operating value from the same camera estate, including queue visibility, safety monitoring, layout intelligence, and cross-store pattern detection.

At an executive level, this value driver matters because shrink and fraud are no longer isolated LP problems. They affect margin, inventory accuracy, associate confidence, customer trust, and store execution. Cisco helps connect the video, network, workflow, security, and analytics layers needed to turn LP from a reactive review process into a more connected operating capability.


### Illustrative example


At a self-checkout area, a camera and POS system identify a repeated mismatch between scanned items and items placed in the bagging area. The system does not prompt a floor associate to confront the customer. Instead, the event is routed to the appropriate team for review with the video clip, transaction data, and store context attached. Over time, the retailer identifies a pattern tied to specific lanes and times of day, adjusts staffing and signage, and improves both shrink control and customer flow.


## Value Driver 2: Intelligent Labor and Service Flow



### What challenge it solves


Store labor is often managed with limited visibility into what is happening in the moment. Managers may know the schedule, but not where service friction is forming. Associates may be busy with tasks while a queue builds at returns. A department may need coverage because customers are dwelling without help. A pickup customer may be waiting while the responsible associate is unaware they have arrived. Fitting rooms, service counters, checkout lanes, pickup areas, and high-consideration departments can all become friction points before leadership sees the problem.

This creates a reactive operating model. Store teams respond after customers are already waiting, after associates are already stretched, or after a service issue has already escalated.

Intelligent labor and service flow gives retailers a way to sense friction earlier and route work more clearly. The objective is not to monitor associates for its own sake. It is to protect scarce labor capacity and make store work easier to coordinate.


### Use cases Cisco enables


- Queue detection: Identify congestion at checkout, returns, service counters, pickup zones, pharmacy, or other high-traffic areas.

- Dynamic associate tasking: Route work to associates based on live store conditions, department activity, and service needs.

- Pickup readiness: Notify associates when pickup or curbside customers are arriving or when order staging requires action.

- Department coverage: Detect high dwell or traffic in departments where customers may need assistance.

- Returns flow management: Identify service desk congestion and trigger staffing or communication workflows.

- Fitting room support: Notify associates when assistance, item retrieval, or room turnover is needed.

- Dwell-based intervention: Identify high-consideration zones where customers may need help deciding.

- Store manager visibility: Give leaders a clearer real-time view of where execution is breaking down.


### Technologies used


Cisco Spaces-style location intelligence helps retailers understand traffic, dwell, and zone-level activity across the store.

Meraki cameras and sensors can support visibility into queues, occupancy, service areas, and traffic patterns.

Cisco wireless and switching provide the connectivity foundation for associate devices, operational systems, and store analytics.

Webex Connect can route operational alerts, task-related messages, and workflow notifications to the right teams.

Webex AI can support prioritization, summarization, and intent classification for service and tasking workflows.

Splunk can correlate operational signals across traffic, staffing, POS, fulfillment, and service systems.

AppDynamics and ThousandEyes help assure applications, APIs, and network paths that support tasking, pickup, and fulfillment flows.


### Business impacts


Intelligent labor and service flow creates value by helping stores respond to friction before customers are already waiting and associates are already overloaded.

- Zebra’s 17th Annual Global Shopper Study coverage found that roughly eight in 10 retail associates are concerned about the lack of technology deployed to spot safety threats or criminal activity, underscoring the need for better frontline visibility.

- Lower abandonment risk when queues, service counters, pickup zones, fitting rooms, and high-intent departments are visible sooner.

- Better associate utilization when traffic, dwell, staffing, fulfillment, and task signals can guide where work is needed most.

- Improved pickup, returns, pharmacy, service desk, and fitting-room flow through earlier detection and clearer task routing.

- Improved employee experience when associates receive clearer direction instead of compensating for disconnected systems.

At an executive level, this value driver matters because labor is one of retail’s scarcest resources. Cisco helps make labor more responsive by connecting location intelligence, cameras, sensors, messaging, tasking, and operational telemetry into a clearer view of store conditions.


### Illustrative example


A general merchandise store sees a returns line building on Saturday afternoon. Instead of waiting for customer complaints, the system detects congestion, correlates it with staffing and transaction volume, and sends a task to a cross-trained associate. A second associate receives a prompt to handle pickup orders separately so the returns line does not block fulfillment. The store improves customer flow without adding labor.


## Value Driver 3: AI-Assisted Workflow Protection



### What challenge it solves


Store associates and contact center agents spend significant time answering questions that are repetitive but still context-dependent. Customers ask where their order is, whether an item is in stock, what the return policy is, whether a pickup is ready, where a product is located, whether a coupon applies, or whether one product works with another.

These questions may seem simple, but they interrupt store work constantly. Every routine question that reaches an associate can pull them away from selling, fulfillment, recovery, shelf work, service, or safety responsibilities. The same pattern appears in the contact center, where routine inquiries consume agent capacity and increase cost-to-serve.

AI-assisted workflow protection reduces that burden by helping customers and associates get routine answers faster while preserving escalation for more complex needs.

The goal is not to keep customers away from people. It is to protect human capacity for moments where human judgment, empathy, physical action, or expert guidance is actually needed.


### Use cases Cisco enables


- Routine customer question automation: Resolve common questions through Webex-powered messaging or AI-assisted flows.

- Associate assist: Provide associates with quick answers from policy, inventory, product, order, or service systems.

- Order status deflection: Automate pickup, delivery, and fulfillment updates before customers call or ask in store.

- Return policy guidance: Provide consistent answers before a customer reaches the service counter.

- Product Q&A support: Help with product details, compatibility, availability, or usage questions.

- Escalation routing: Move complex issues to store teams or contact center agents with context attached.

- Post-interaction summaries: Reduce manual notes and improve continuity across store, digital, and contact center interactions.

- Service recovery triggers: Detect failed or delayed journeys and launch proactive outreach.


### Technologies used


Webex Connect orchestrates messaging and workflow automation across SMS, RCS, WhatsApp, app chat, web chat, and other channels.

Webex AI supports conversational assistance, intent detection, summarization, guided next steps, and automation of routine service tasks.

Webex Contact Center handles human escalation and gives agents better context when automation is not enough.

Cisco networking supports reliable in-store and cloud connectivity for customer, associate, and service workflows.

AppDynamics and ThousandEyes help assure the digital systems that answer customer questions, retrieve order context, or support associate tools.

Splunk can correlate service events, application telemetry, messaging flows, and operational signals.


### Business impacts


AI-assisted workflow protection creates value by reducing the routine questions and disconnected tasks that pull associates away from service, fulfillment, recovery, and selling.

- Directional service benchmarks often use roughly $1.84 for self-service versus $13.50 for assisted support, showing why routine containment can be a meaningful capacity lever.

- Zendesk / Forrester TEI evidence supports the broader impact of AI-assisted support on contact-rate reduction, automated resolution, handle time, and agent productivity.

- Fewer associate interruptions for order status, item location, return policy, product availability, coupon, and product-comparison questions.

- Improved service consistency when policy, inventory, order, loyalty, and product information can be surfaced through AI-assisted workflows.

- More human capacity for moments that require judgment, empathy, physical action, or expert guidance.

At an executive level, this value driver matters because better EX is not only about adding tools. It is about protecting associate attention. Cisco helps connect AI, Webex workflows, contact center, networking, and observability so routine work can be handled more consistently and complex work can reach the right person with context.


### Illustrative example


An apparel shopper asks whether an item is available in another size. Instead of searching manually across systems, the associate uses an AI-assisted workflow on a handheld. The system checks local inventory, nearby store availability, and ship-to-home options, then offers to reserve the item or send a purchase link. If the customer asks about returns, the interaction can move to a Webex-powered messaging flow. The associate stays focused on service rather than system navigation.


## Value Driver 4: Task Automation and ESLs



### What challenge it solves


Retail stores still rely on many manual execution tasks. Price changes, promotion updates, shelf checks, signage replacement, compliance audits, item lookups, and exception handling consume associate time every week. These tasks are repetitive, error-prone, and often compete with customer service and fulfillment work.

Pricing and promotion errors also affect customer trust. If the shelf price does not match the register price, the customer sees the retailer as inconsistent. If a promotion remains posted after inventory is gone, the store creates frustration. If associates spend hours replacing labels, they are not available for higher-value work.

Task automation and electronic shelf labels can help reduce that burden. The broader opportunity is not only to digitize price tags. It is to connect shelf execution to pricing, promotion, inventory, media, associate workflows, and customer information.


### Use cases Cisco enables


- Automated price updates: Reduce manual label changes and improve shelf-to-POS price consistency.

- Promotion execution: Update shelf-edge content in sync with campaigns, offers, and store events.

- Inventory-aware shelf messaging: Support local availability, substitutions, order options, or product alternatives.

- Shelf-edge product education: Use ESLs or labels to launch product information, comparisons, recipes, reviews, or scan-to-learn flows.

- Compliance support: Reduce mismatches between shelf price, advertised price, and POS price.

- Associate tasking: Route exceptions when updates fail or shelves need attention.

- Retail media activation: Support supplier-funded shelf-edge content where appropriate and measurable.

- Operational visibility: Monitor update success, device health, and exception patterns across stores.


### Technologies used


Cisco Meraki and Catalyst networking connect store systems, ESL gateways, associate devices, and supporting applications.

PoE switching can support connected infrastructure where relevant, including access points, gateways, sensors, cameras, and signage.

Meraki access point and ESL integration patterns can reduce separate gateway complexity where supported and approved.

Webex Connect can support customer or associate workflows triggered by shelf-edge interactions, product questions, or service needs.

Splunk can correlate ESL updates, pricing events, device status, and operational telemetry.

AppDynamics and ThousandEyes help assure supporting applications, APIs, and cloud services that update pricing or content.

Cisco security and segmentation help protect ESL infrastructure and related store systems.


### Business impacts


Task automation and ESLs create value by reducing repetitive shelf-execution work and improving the accuracy of pricing, promotion, and product information in the store.

- Walmart has announced a rollout of digital shelf labels across its U.S. stores by the end of 2026, highlighting how quickly shelf execution is becoming a digital operating layer.

- Reduced manual label-change labor when price and promotion updates can be executed digitally instead of through paper-tag replacement.

- Improved shelf-to-POS consistency, reducing customer disputes and pricing-error friction.

- Faster promotion execution and greater flexibility for local inventory, daypart, campaign, or category changes.

- A stronger foundation for shelf-edge product education, dynamic merchandising, and in-store retail media.

At an executive level, this value driver matters because store teams spend too much time on repetitive work that can be automated or exception-managed. Cisco helps connect ESL infrastructure, store networking, PoE, application assurance, security, and telemetry so shelf execution becomes more accurate, visible, and adaptable.


### Illustrative example


A home improvement retailer needs to update prices across thousands of SKUs during a weekend promotion. Instead of associates spending hours replacing paper tags, ESLs update centrally. Exceptions are flagged only where a shelf update fails. Digital shelf content also supports product comparisons and accessory suggestions for high-consideration items, giving the retailer both labor efficiency and better customer information at the shelf.


## Value Driver 5: Connected Facilities and Smart Lighting



### What challenge it solves


Facilities systems are often managed separately from store operations. Lighting, HVAC, refrigeration, occupancy, energy management, and building controls may operate on independent systems with limited connection to customer traffic, associate workflows, safety needs, merchandising, or IT visibility.

That separation creates missed opportunities. Lighting may remain at full output in low-traffic zones. Occupancy data may not inform operations. Facilities issues may not be visible until they affect customers or associates. Energy savings may be pursued separately from safety, merchandising, or service goals. Sensors may exist but not contribute to broader store intelligence.

Connected facilities and smart lighting create an opportunity to make building systems part of the store operating model.

Smart lighting should not be treated as a standalone pillar. It is one practical example of how store infrastructure can become connected, controllable, measurable, and operationally useful.


### Use cases Cisco enables


- Occupancy-based lighting: Adjust lighting based on traffic, staffing, store hours, or zone activity.

- Daylight harvesting: Reduce energy use when natural light is sufficient.

- Task lighting: Support stocking, service counters, pickup areas, fitting rooms, or high-value departments.

- Safety support: Improve visibility in parking, backroom, loading, or restricted areas.

- Merchandising emphasis: Highlight promotional zones, endcaps, or product displays.

- Facilities automation: Coordinate lighting with HVAC, occupancy, and store schedules.

- Telemetry integration: Feed environmental and usage data into operational dashboards.

- After-hours monitoring: Detect unusual occupancy, lighting, or facility activity outside expected patterns.


### Technologies used


Cisco Catalyst PoE and PoE++ switching can power and connect lighting and sensor infrastructure in suitable deployments.

Cisco Meraki networking supports cloud-managed store connectivity and operational visibility.

Sensors and IoT systems capture occupancy, environmental, and facility signals.

Splunk can correlate facilities telemetry with store operations, energy data, security events, and occupancy patterns.

Cisco security and segmentation help protect facilities and IoT systems from broader network exposure.

AppDynamics and ThousandEyes can help assure supporting applications and cloud services where relevant.


### Business impacts


Connected facilities and smart lighting create value by turning building systems into part of the store operating model instead of treating them as isolated facilities infrastructure.

- A Lawrence Berkeley National Laboratory meta-analysis of lighting controls reviewed 240 energy-savings estimates across 88 papers and case studies, reinforcing lighting controls as a well-studied savings lever.

- Retail TEV planning commonly isolates lighting-related energy before applying savings assumptions, because lighting can represent a meaningful controllable share of store utility spend.

- Lower energy consumption when lighting responds to traffic patterns, operating hours, daylight, and zone activity.

- Improved visibility and safety in pickup areas, parking lots, back rooms, loading zones, stocking areas, and service counters.

- More useful occupancy and environmental data for operations, facilities, security, and sustainability reporting.

At an executive level, this value driver matters because facilities systems can support more than cost control. When connected properly, they contribute to safety, sustainability, operational visibility, and the broader store intelligence layer. Cisco helps provide the PoE, connectivity, segmentation, and telemetry foundation for that shift.


### Illustrative example


A large-format retailer uses connected lighting to dim low-traffic zones during late evening hours while keeping pickup, checkout, and restocking areas fully lit. Occupancy sensors feed facilities data into an operations dashboard. The retailer saves energy while improving visibility where associates are actively working.


## Value Driver 6: Connected Device Reliability



### What challenge it solves


Store operations depend on connected devices: handhelds, scanners, tablets, POS terminals, self-checkout lanes, cameras, sensors, digital signage, ESLs, kiosks, and associate communication tools. When these devices fail or perform poorly, associates lose time, customers wait, and IT teams struggle to identify root cause.

The problem is often difficult to diagnose because the symptom appears in one place while the cause sits somewhere else. A handheld may appear broken when the real issue is Wi-Fi interference. A pickup workflow may slow down because of an API dependency. A digital sign may fail because of a content system issue. A POS delay may involve a WAN path, payment service, or local network condition. A camera outage may involve power, connectivity, policy, or cloud access.

Without connected observability, troubleshooting becomes reactive. Store teams report problems. IT investigates. Vendors are contacted. Time passes. The customer and associate experience suffers.

Connected device reliability gives retailers a way to see and resolve issues faster.


### Use cases Cisco enables


- Handheld performance assurance: Detect Wi-Fi, application, API, or device issues affecting associate tools.

- POS and checkout resilience: Monitor connectivity and dependencies that support payment and transaction workflows.

- Camera and sensor uptime: Ensure operational visibility systems remain available.

- Digital signage assurance: Detect failed content delivery, connectivity issues, or device outages.

- ESL update monitoring: Identify update failures and supporting network or application issues.

- Incident correlation: Connect device symptoms to application, network, cloud, security, or API root causes.

- Remote troubleshooting: Reduce truck rolls and store visits by diagnosing issues centrally.

- Priority-based remediation: Focus IT response on issues affecting critical workflows.


### Technologies used


Cisco wireless, switching, and SD-WAN provide reliable store connectivity for devices and applications.

Meraki cloud-managed infrastructure supports visibility and remote management across distributed store environments.

AppDynamics monitors application performance and business transactions.

ThousandEyes provides visibility into WAN, internet, SaaS, and cloud paths.

Splunk correlates telemetry across devices, applications, networks, security, and incidents.

Cisco security capabilities help protect devices, enforce policy, and identify suspicious behavior.


### Business impacts


Connected device reliability creates value by reducing the operational drag caused when handhelds, scanners, POS, cameras, signage, ESLs, kiosks, and associate tools fail or perform poorly.

- Cisco’s Meraki economic-impact materials cite a 40% reduction in network downtime and a 90% reduction in time spent on configuration changes in a commissioned Forrester TEI study.

- Fewer store visits and truck rolls when IT can diagnose network, device, application, and cloud-path issues remotely.

- Reduced MTTR when device symptoms can be correlated with Wi-Fi, WAN, API, SaaS, power, security, or application root causes.

- Improved associate productivity when critical tools stay available for pickup, inventory, checkout, tasking, and customer service.

- Clearer IT prioritization because incidents can be tied to affected stores, workflows, and business impact.

At an executive level, this value driver matters because device reliability is now business reliability. Cisco helps connect wireless, switching, SD-WAN, cloud management, AppDynamics, ThousandEyes, Splunk, and security telemetry so IT can move from reactive troubleshooting to proactive assurance.


### Illustrative example


A grocery chain sees intermittent handheld failures in multiple stores. Associates think the devices are broken, but observability shows that the issue is tied to a specific SaaS API path during peak inventory updates. IT reroutes traffic and resolves the issue centrally, avoiding unnecessary device replacements and store visits while restoring associate productivity.


# 8. Priority 3 - IT Resilience, Trust, and Protection


In the connected store, IT resilience becomes business resilience.

A network outage is no longer just a technical event. It can stop payments, disrupt pickup, break loyalty redemption, disable digital signage, interrupt cameras, slow handhelds, delay associate workflows, and create customer frustration. An application slowdown may not appear as a full outage, but it can still damage conversion, productivity, and trust. A cyber incident can affect customer data, store systems, vendor access, payment environments, and daily operations.

AI-native retail increases this dependency. More devices, more applications, more APIs, more cloud services, more edge workloads, more customer data, and more automated workflows mean the store needs a stronger foundation. Every new connected capability can create value, but every new connection also adds complexity and risk if it is not secured, observed, and governed.

This is why IT can no longer be positioned as a back-office support function in the retail technology story. IT is the foundation that determines whether CX and EX improvements can scale. The same infrastructure that supports customer engagement also supports associate productivity, loss prevention, digital signage, smart lighting, fulfillment, contact center, observability, and cyber resilience.

For retailers, the strategic question is not simply, “Can the network stay online?” It is, “Can the technology foundation support the connected store without increasing fragility?”

Cisco’s role in this priority is to help retailers build the secure, observable, AI-ready infrastructure required to support modern stores at scale. That includes connectivity, identity, segmentation, threat detection, application assurance, network visibility, cloud path monitoring, telemetry correlation, and edge/cloud readiness.

Within this priority, five value drivers matter most:


|Value driver|What it does|
|---|---|
|Secure Connected Store Foundation|Provides reliable connectivity for stores, devices, applications, and cloud services|
|Zero Trust and Store Segmentation|Protects users, devices, applications, IoT, guest Wi-Fi, and payment-related systems|
|Observability as a Business Capability|Connects technical performance to customer, associate, and store workflow impact|
|Cyber Resilience and Risk Reduction|Reduces exposure to ransomware, account takeover, credential theft, vendor risk, and shadow AI|
|AI-Ready Edge and Cloud Infrastructure|Supports distributed AI workloads, cameras, sensors, APIs, and real-time store intelligence|



## Value Driver 1: Secure Connected Store Foundation



### What challenge it solves


Retail stores depend on reliable connectivity, but many store networks were designed before the current density of cloud applications, mobile devices, smart cameras, sensors, digital signage, electronic shelf labels, self-checkout, and AI-enabled workflows.

Legacy store networks can become bottlenecks when retailers add more devices and more cloud dependencies. What used to be a relatively contained store environment is now a distributed technology node. A single store may support guest Wi-Fi, associate handhelds, POS terminals, payment devices, self-checkout, digital signage, cameras, sensors, smart lighting, ESL infrastructure, kiosks, contact center handoffs, pickup workflows, and multiple cloud applications.

If the foundation is weak, every connected use case becomes fragile.

The connected store requires a network foundation that is reliable enough for daily operations, flexible enough for new devices, secure enough for sensitive systems, and observable enough for IT to understand what is happening across hundreds or thousands of locations.


### Use cases Cisco enables


- Store Wi-Fi modernization: Support associate devices, customer engagement, guest Wi-Fi, and operational systems with reliable connectivity.

- Cloud-managed store operations: Give IT teams centralized visibility and management across distributed locations.

- SD-WAN resilience: Improve store connectivity to cloud, SaaS, and enterprise systems.

- PoE device support: Power and connect cameras, access points, sensors, signage, ESL gateways, and selected smart lighting deployments.

- Critical workflow prioritization: Protect traffic for POS, payment, pickup, associate tools, and other high-priority store systems.

- Device visibility: Understand what is connected across store environments and identify unmanaged or unexpected devices.

- Remote troubleshooting: Reduce dependence on store visits by diagnosing network and device issues centrally.

- New use case readiness: Create a foundation that can support future connected-store deployments without rebuilding the store network each time.


### Technologies used


Cisco Catalyst and Meraki wireless provide high-performance store connectivity for associates, customers, devices, and operational systems.

Cisco switching supports access, aggregation, PoE, and wired device connectivity across the store.

Cisco SD-WAN improves store-to-cloud and store-to-enterprise resilience by optimizing connectivity across WAN, internet, and cloud paths.

PoE and PoE++ switching can power and connect access points, cameras, sensors, signage, and selected smart lighting or edge infrastructure.

Meraki cloud management provides centralized visibility and simplified operations across distributed store environments.

ThousandEyes provides visibility into internet, WAN, cloud, and SaaS paths that affect store applications and customer journeys.

Splunk can correlate infrastructure, security, application, and operational telemetry to help IT teams see patterns across stores.


### Business impacts


A secure connected store foundation creates value by making store infrastructure reliable enough, visible enough, and secure enough to support the next wave of connected retail use cases.

- Cisco’s Meraki economic-impact materials report $7.3M in total benefits over three years, $1.7M in savings, and a 90% reduction in time spent on configuration changes in a commissioned Forrester TEI study.

- Lower support burden when cloud-managed networking gives IT teams centralized visibility across distributed stores.

- Fewer unnecessary truck rolls when network and device issues can be diagnosed remotely.

- Improved readiness for cameras, sensors, signage, ESLs, smart lighting, associate devices, mobile POS, and AI-enabled workflows.

- Better continuity for critical store workflows such as POS, payment, pickup, loyalty, inventory, and associate tasking.

At an executive level, this value driver matters because the network is no longer only infrastructure. It is the execution foundation for CX, EX, LP, retail media, facilities, security, and AI. Cisco helps retailers modernize that foundation without treating every new use case as a separate store rebuild.


### Illustrative example


A specialty retailer rolls out mobile POS and clienteling tablets across stores. In locations with older Wi-Fi, associates experience intermittent disconnects that slow checkout and frustrate customers. A modernized Cisco foundation improves device performance, gives IT visibility into store conditions, and supports future digital signage, smart camera, and ESL use cases without requiring a separate infrastructure rebuild.


## Value Driver 2: Zero Trust and Store Segmentation



### What challenge it solves


Modern stores contain many types of connected users, devices, and systems: POS, payment devices, guest Wi-Fi, associate handhelds, cameras, digital signage, ESL gateways, HVAC, smart lighting, sensors, kiosks, vendors, support teams, cloud applications, and customer engagement platforms.

If these systems operate on overly flat networks or weak access policies, a compromise in one area can create risk across the store. A vendor account may have more access than it needs. A digital signage player may sit too close to sensitive systems. A guest network may not be adequately separated. A camera or IoT device may become a path into broader infrastructure. A compromised credential may expose more systems than intended.

The connected store expands the attack surface. Zero trust and segmentation help retailers expand connected capabilities without expanding risk at the same pace.

The goal is not to make store operations harder. The goal is to allow more devices, applications, vendors, and AI-enabled workflows to operate safely.


### Use cases Cisco enables


- Guest and corporate network separation: Prevent guest access from touching store systems.

- IoT segmentation: Isolate cameras, signage, ESLs, lighting, sensors, HVAC, and other operational devices.

- Payment environment protection: Restrict access to POS and payment-related systems.

- Context-aware access: Validate users, devices, posture, location, and risk before granting access.

- Vendor access control: Limit third-party access to only the systems, time windows, and functions required.

- Policy consistency: Apply access and segmentation policies across distributed stores.

- Threat containment: Reduce lateral movement and limit blast radius when a device, credential, or application is compromised.

- AI and data governance support: Control access to systems and data that AI workflows may use.


### Technologies used


Cisco ISE supports identity-based access control, device profiling, and segmentation policies across store networks.

Cisco Duo supports multi-factor authentication and device trust for users accessing applications and systems.

Cisco Secure Access helps secure access to private applications, internet destinations, and cloud services.

Cisco networking and segmentation capabilities enforce access boundaries across wired, wireless, guest, corporate, IoT, and operational systems.

Cisco XDR supports detection and response across security telemetry.

Cisco Umbrella / DNS-layer security can help block malicious destinations and reduce exposure to internet-based threats.

Splunk can correlate security, identity, network, and operational events to support investigation and response.

Meraki and Catalyst infrastructure support policy enforcement across distributed store environments.


### Business impacts


Zero trust and store segmentation create value by allowing retailers to connect more devices, vendors, applications, and workflows without expanding risk at the same pace.

- IBM’s 2025 breach-cost research reports a $4.44M global average cost of a data breach, reinforcing why segmentation, identity, and access controls matter as stores connect more systems.

- Reduced blast radius when POS, payment, guest Wi-Fi, cameras, signage, ESLs, HVAC, sensors, and associate devices are segmented appropriately.

- Stronger protection for customer data, payment environments, loyalty systems, and sensitive store applications.

- Lower vendor-access risk when third parties are limited to the systems, time windows, and functions they actually need.

- Safer expansion of IoT, AI, signage, cameras, and facilities systems because access is governed by context and policy.

At an executive level, this value driver matters because the connected store expands the attack surface. Cisco helps retailers enforce trust across users, devices, applications, IoT, vendors, and data flows so innovation can scale without leaving store systems overly exposed.


### Illustrative example


A retailer connects new digital signage, ESL gateways, and smart lighting systems across hundreds of stores. Instead of placing all devices on a flat store network, IT segments each device category and restricts access to required services only. If one signage device is compromised, segmentation helps prevent access to payment systems, associate devices, customer data, and enterprise applications.


## Value Driver 3: Observability as a Business Capability



### What challenge it solves


In a connected store, customer and associate experiences depend on applications, APIs, networks, cloud services, ISPs, devices, and third-party platforms. Traditional monitoring may show that systems are technically online, but it may not reveal why a pickup workflow is slow, why an app coupon fails, why a handheld drops, why a digital sign is blank, or why checkout latency is increasing.

Retailers need observability that connects technical performance to business journey impact.

That means moving beyond the question, “Is the system up?” to more useful questions:

Can the customer redeem a coupon?

Can the associate complete a pickup?

Can the POS process payment?

Can the app load the product page?

Can the ESL update complete?

Can digital signage retrieve content?

Can the contact center receive context?

Can the camera system upload an event?

Can the order management API respond?

Which stores, customers, associates, or workflows are affected?

Without this visibility, IT teams troubleshoot in the dark, store teams absorb the friction, and customers feel the impact before the enterprise understands the root cause.


### Use cases Cisco enables


- Checkout journey assurance: Monitor payment, POS, network, and API dependencies that affect checkout.

- Pickup and fulfillment monitoring: Track app, OMS, messaging, and store workflow performance.

- Digital experience assurance: Monitor app, site, loyalty, offer redemption, and product information performance.

- Store device incident correlation: Connect handheld, signage, camera, or ESL failures to root causes.

- SaaS and internet visibility: Understand third-party path issues beyond the store network.

- Incident prioritization: Prioritize issues based on customer, associate, and business workflow impact.

- Executive reporting: Translate technical events into business impact language.

- Proactive remediation: Identify degraded performance before it becomes a widespread customer or store issue.


### Technologies used


AppDynamics monitors application performance, APIs, and business transactions that support customer and associate journeys.

ThousandEyes provides visibility into WAN, internet, SaaS, cloud, and third-party paths that affect store and digital experiences.

Splunk correlates telemetry across applications, networks, security, devices, and operational systems.

Cisco networking telemetry provides device and network performance context.

Meraki cloud management supports visibility into distributed store environments.

Webex platforms can provide interaction, contact center, and communication context where relevant.

Cisco security telemetry can add identity, access, and threat context to operational incidents.


### Business impacts


Observability creates value by helping retailers connect technical performance to the customer, associate, and store workflows that depend on it.

- Cisco Full-Stack Observability TEI materials report a 60% reduction in the average length of a major outage by Year 3 and a 90% reduction in time spent on war-room triage and resolution.

- Faster diagnosis when application, API, SaaS, internet path, network, device, and security telemetry can be correlated.

- Reduced MTTR and support effort by showing which systems, stores, and journeys are affected by degradation.

- Better protection for checkout, pickup, loyalty redemption, digital signage, contact center, ESL updates, and associate tools.

- Improved executive visibility because IT incidents can be explained in terms of business workflow impact, not only technical status.

At an executive level, this value driver matters because “green dashboards” are not enough when customers and associates are still feeling friction. Cisco helps retailers observe the full path from store to application to cloud so IT can detect, diagnose, prioritize, and resolve issues faster.


### Illustrative example


A retailer sees a spike in pickup complaints in one region. Store teams think the issue is staffing. IT initially sees no major outage. Observability shows that a third-party API used by the order-status service is intermittently timing out through a specific internet path. ThousandEyes identifies the path issue, AppDynamics shows the affected transaction, and Splunk correlates complaint volume. The retailer resolves the issue before it becomes a national service problem.


## Value Driver 4: Cyber Resilience and Risk Reduction



### What challenge it solves


Retailers are attractive targets because they operate payment systems, customer accounts, loyalty programs, gift cards, ecommerce platforms, vendor integrations, supply-chain connections, and distributed store networks. Ransomware, credential theft, account takeover, API abuse, supply-chain compromise, IoT exposure, phishing, and shadow AI can all create operational and financial risk.

The connected store increases both value and exposure. More devices, more applications, more vendors, more cloud services, more data flows, and more AI-enabled workflows mean more places where attackers, misconfigurations, or weak policies can create harm.

Cyber resilience is therefore not only a security requirement. It is a business continuity requirement.

A cyber incident can disrupt stores, expose customer data, interrupt fulfillment, damage loyalty, create regulatory exposure, and slow operations. Retailers need to detect, contain, investigate, and recover quickly while keeping critical store operations moving.


### Use cases Cisco enables


- Credential and access protection: Reduce risk from compromised accounts, weak authentication, or unmanaged access.

- Threat detection and response: Detect suspicious activity across endpoints, network, cloud, identity, and applications.

- Store segmentation: Limit blast radius from compromised devices or systems.

- Shadow AI and data governance support: Reduce unsanctioned data exposure and improve access controls where applicable.

- Vendor risk containment: Limit third-party access to specific systems, time windows, and purposes.

- Security telemetry correlation: Connect security events with operational, identity, network, and application context.

- Incident response support: Improve investigation speed, containment, and executive visibility.

- Resilience planning: Protect critical workflows such as POS, fulfillment, contact center, and store operations.


### Technologies used


Cisco Duo provides multi-factor authentication and device trust to reduce credential-related risk.

Cisco Secure Access supports secure access to applications, internet destinations, and cloud services.

Cisco ISE helps enforce identity-based access and segmentation policies.

Cisco XDR supports detection and response across multiple security telemetry sources.

Cisco Umbrella / DNS-layer security helps block malicious destinations and reduce internet-based exposure.

Splunk correlates security, IT, identity, and operational events for investigation and response.

Cisco networking and segmentation provide enforcement points for access control and containment.


### Business impacts


Cyber resilience creates value by protecting the connected store from disruptions that can affect operations, customer trust, payment environments, loyalty accounts, vendors, and data flows.

- Sophos retail ransomware research reported a mean recovery cost of $2.73M in 2024, while IBM’s 2025 breach-cost research reports a $4.44M global average data-breach cost.

- Retailers face persistent exposure from ransomware, credential theft, account takeover, phishing, API abuse, vendor access, and IoT compromise.

- Reduced likelihood and impact of incidents when identity, device trust, secure access, segmentation, threat detection, and telemetry correlation work together.

- Faster containment and investigation when security signals can be tied to users, devices, applications, stores, and operational workflows.

- Protected customer trust when customer data, loyalty accounts, payment systems, and sensitive applications are governed more consistently.

At an executive level, this value driver matters because cybersecurity in retail is now a business continuity issue. Cisco helps reduce exposure, contain incidents, and improve response by connecting identity, segmentation, secure access, detection, DNS-layer protection, Splunk, and network enforcement.


### Illustrative example


A retailer detects suspicious login behavior from a third-party maintenance account. Instead of broad access, the account is governed through identity controls, device trust, and segmented network policy. Security telemetry is correlated in Splunk, and the access attempt is contained before it reaches store systems. The incident becomes a manageable security event rather than an operational disruption.


## Value Driver 5: AI-Ready Edge and Cloud Infrastructure



### What challenge it solves


AI-native retail use cases increasingly depend on data and decisioning across the edge, cloud, and enterprise systems. Computer vision, queue detection, smart lighting, digital signage, product assistants, associate workflows, dynamic offers, inventory intelligence, and real-time tasking may all require low latency, local processing, secure APIs, resilient connectivity, and governed data flows.

Retailers that add AI workloads on top of brittle infrastructure risk creating new silos. One team deploys a computer vision pilot. Another deploys a digital signage platform. Another tests an associate assistant. Another adds sensors. Another adds customer messaging. If each project creates its own architecture, the store becomes harder to secure, harder to observe, and harder to scale.

AI-ready infrastructure gives retailers a reusable foundation for current and future use cases.

The value is not only the first AI deployment. It is the ability to deploy the next one faster, more securely, and with less rework.


### Use cases Cisco enables


- Edge video analytics: Support local processing for cameras and computer vision use cases.

- Real-time store intelligence: Process traffic, queue, sensor, device, and operational signals.

- AI-assisted customer journeys: Connect product, loyalty, inventory, messaging, and service systems.

- AI-assisted associate workflows: Support handheld, tasking, and associate-assist experiences.

- Dynamic content decisioning: Connect signage and merchandising to real-time store conditions.

- Hybrid cloud connectivity: Support secure data movement between store, cloud, and enterprise systems.

- Future AI workload readiness: Prepare stores for new AI use cases without rebuilding the foundation.

- Governed data flows: Help ensure AI systems access the right data with the right controls.


### Technologies used


Cisco networking, switching, and SD-WAN provide connectivity for edge, store, cloud, and enterprise workloads.

PoE infrastructure supports connected edge devices such as cameras, access points, sensors, signage, and selected smart infrastructure.

Meraki cameras and sensors provide physical-world data sources for store intelligence.

AppDynamics and ThousandEyes assure applications, APIs, cloud paths, and network dependencies that AI workflows rely on.

Splunk correlates data and telemetry across AI-related workflows, applications, devices, and operations.

Cisco security capabilities protect APIs, devices, users, data flows, and AI-adjacent systems.

Webex AI, Webex Connect, and Webex Contact Center support AI-enabled customer engagement, service workflows, and escalation.


### Business impacts


AI-ready edge and cloud infrastructure creates value by giving retailers a reusable foundation for current and future AI-enabled store use cases.

- McKinsey has estimated that generative AI could unlock $240B to $390B in value for retailers, and broader retail / CPG estimates reach $400B to $660B annually when use cases are fully implemented.

- Faster rollout of use cases such as computer vision, queue detection, smart lighting, dynamic signage, associate assist, product Q&A, and operational automation.

- Reduced infrastructure rework when new AI pilots can build on a common connectivity, security, observability, and data-flow foundation.

- Improved governance because AI workflows can access store, edge, cloud, application, and customer data through controlled paths.

- Greater strategic flexibility as retail formats, AI capabilities, device ecosystems, and customer expectations change.

At an executive level, this value driver matters because AI value does not come from isolated pilots. It comes from the ability to connect intelligence to execution repeatedly. Cisco helps create the edge, cloud, security, observability, and workflow foundation that lets retailers scale AI use cases without rebuilding the store architecture every time.


### Illustrative example


A retailer wants to expand computer vision from loss prevention to queue detection, inventory visibility, and safety analytics. Rather than deploying separate infrastructure for each pilot, the retailer uses a connected edge and cloud foundation that supports camera data, secure segmentation, application monitoring, and workflow integration. Each new use case builds on the same foundation instead of creating another silo.


# 9. Cisco Connected Retail Architecture


The connected store operating model requires an architecture that can support business outcomes, not just technology deployment. Retailers do not need a product catalog disguised as a strategy. They need a clear view of the layers that must work together for the store to become more connected, intelligent, secure, and measurable.

Cisco’s role is strongest at the intersections: where devices connect to networks, where applications depend on cloud services, where customer engagement depends on messaging and contact center, where cameras and sensors create operational signals, where security policy protects distributed systems, and where observability connects technical performance to business impact.

The architecture should be understood as a shared operating foundation rather than a stack of isolated layers. The same connectivity layer that supports POS can also support cameras, associate devices, digital signage, ESLs, smart lighting, and customer engagement. The same observability layer that helps IT troubleshoot application issues can also protect pickup, loyalty, and checkout journeys. The same security controls that protect payment environments can also make it safer to add IoT, signage, and edge devices. The same Webex engagement layer that supports service recovery can also support product-originated conversations, pickup coordination, and contact center escalation.


|Layer|Role|Representative Cisco capabilities|
|---|---|---|
|Connectivity and transit|Connect stores, devices, applications, cloud services, and critical workflows.|Catalyst, Meraki, switching, SD-WAN, cloud-managed networking|
|Power and device infrastructure|Power and manage cameras, access points, sensors, signage, gateways, and edge devices.|PoE / PoE++, Meraki access infrastructure, policy enforcement|
|Store intelligence|Convert physical activity into operational context for traffic, queues, safety, shrink, merchandising, and facilities.|Cisco Spaces-style location intelligence, Meraki cameras/sensors, Splunk|
|Customer engagement and CPaaS|Turn store and customer context into timely communication and escalation.|Webex Connect, Webex AI, Webex Contact Center|
|Observability and assurance|Connect technical performance to customer, associate, and store workflow impact.|AppDynamics, ThousandEyes, Splunk, network telemetry|
|Security and trust|Protect users, devices, applications, IoT, vendors, and AI-adjacent data flows.|Duo, ISE, Secure Access, XDR, Umbrella, segmentation, Splunk|
|Edge, cloud, and AI readiness|Support distributed AI workloads and governed data movement across store, edge, cloud, and enterprise systems.|SD-WAN, secure store-to-cloud, Meraki cameras/sensors, AppDynamics, ThousandEyes, Splunk|


The connected retail architecture gives retailers a way to scale innovation without recreating fragmentation each time a new use case appears. Detailed layer-by-layer architecture planning should follow in technical workshops or technical workshops or enablement materials; the whitepaper’s purpose is to show how those layers work together to support the business model.


# 10. From Strategy to Measurable Value


A connected store strategy is only useful if it can be translated into measurable business value.


### Connecting the whitepaper to TEV


This whitepaper defines the operating model and the value drivers. The companion TEV page translates those drivers into scenario-specific estimates, using customer inputs, benchmark ranges, deployment assumptions, and overlap controls. Keeping that separation lets this paper remain strategic while still giving Cisco teams, partners, and retail stakeholders a credible path from narrative to numbers.

For retailers, the opportunity is broad. A more connected store can improve customer engagement, reduce friction in omnichannel journeys, protect associate capacity, improve task execution, reduce shrink, strengthen safety, lower energy consumption, reduce support burden, improve uptime, protect digital experiences, and create new opportunities for retail media and first-party customer insight.

But the value story must be disciplined. Connected-store investments often influence multiple parts of the business at once. The same camera estate that supports loss prevention can also improve queue visibility and safety. The same messaging platform that supports pickup updates can also improve service recovery, contact deflection, loyalty engagement, and product-originated conversations. The same network foundation that supports POS can also support cameras, sensors, signage, ESLs, smart lighting, associate devices, and AI workloads.

That overlap is a strength, but it also requires careful measurement. A credible value case should distinguish between direct operating impact, strategic upside, and protected value.


|Value type|How to treat it|
|---|---|
|Direct operating impact|Measurable improvements such as reduced manual work, lower support burden, fewer truck rolls, reduced energy usage, lower contact volume, faster incident resolution, shrink reduction, or margin-converted revenue uplift.|
|Strategic upside|Capabilities that create future business potential, such as first-party customer context, product-intent signals, retail media capacity, AI-ready infrastructure, and more measurable store engagement.|
|Protected value|Revenue, productivity, customer trust, and operating continuity that the connected-store foundation helps preserve by reducing downtime, improving cyber resilience, and making critical workflows more observable.|


This distinction matters. A retailer should not treat every enabled capability as immediate profit. Revenue opportunities should be converted using appropriate margin assumptions. Productivity improvements should distinguish between hard savings and reallocated capacity. Resilience and cybersecurity value should be framed as protected value, not new revenue. Retail media and first-party data opportunities should be tied to actual monetization and governance maturity.

Cisco Retail 2.0 should therefore move from strategy to value in three steps. First, identify the connected-store priorities that matter most to the retailer. Second, map those priorities to the store systems, workflows, and infrastructure required to execute them. Third, develop a customer-specific value model that separates direct impact, strategic upside, and protected value while using conservative assumptions and clear overlap controls.

The result is not a generic ROI claim. It is a structured value conversation. For the CIO, it shows how infrastructure modernization supports business outcomes rather than only technical refresh. For the COO, it connects store visibility and workflow orchestration to execution. For the CFO, it separates hard savings from strategic upside and protected value. For the CISO, it frames security as business continuity. For the Chief Customer Officer, it connects digital engagement to physical journey improvement.

The strategy defines what the store needs to become. The value model defines what that transformation is worth.


# 11. Summary: The AI-Native Retail Mandate


Retail’s next transformation will not be defined by isolated AI tools. It will be defined by operating models that make AI useful in the real world. That real world is the store.

The store is where customers decide, associates execute, inventory moves, media activates, services are delivered, losses occur, devices connect, and brand promises are kept or broken. It is also where fragmentation becomes visible first.

The AI-native era raises the standard. Retailers cannot simply add intelligence on top of disconnected operations and expect scaled value. AI requires context. Context requires connection. Connection requires secure, resilient, observable infrastructure and workflows that can act on what the business learns.

For customers, the connected store means journeys that feel more relevant, more intuitive, and easier to complete. For associates, it means clearer context, fewer avoidable interruptions, more reliable tools, safer workflows, and less time spent compensating for disconnected systems. For store leaders, it means better visibility into queues, tasks, fulfillment, shrink, safety, inventory, merchandising, and service flow. For IT and security leaders, it means a more observable, governable, resilient environment for the applications, devices, identities, APIs, cloud services, and AI workloads that now shape retail execution.

Cisco’s role is to help retailers build that foundation. Secure networking, Wi-Fi, switching, SD-WAN, PoE, smart cameras, sensors, ESL infrastructure, Webex Connect, Webex AI, Webex Contact Center, AppDynamics, ThousandEyes, Splunk, and Cisco security capabilities can help connect the physical and digital business into one operating environment.

The point is not to make the store feel more technical. The point is to make it work better. The AI-native retailer will not be the one with the most pilots, dashboards, or disconnected digital tools. It will be the one that can connect intelligence to execution across the store estate.

The connected store is how retailers get there.

---

# 5. TEV Landing Page Copy

Cisco Retail 2.0

Total Economic Value Study

A Business Value Framework for the AI-Native Connected Store

Page copy draft for webapp buildout | v1 | May 2026


|Purpose of this page This TEV page translates the Cisco Retail 2.0 connected store operating model into a benchmark-informed business value framework. It is designed for a webapp landing page, not as the detailed calculation engine. The companion calculator carries formulas, source registry, scenario inputs, deployment assumptions, and overlap controls.|
|---|


SECTION 01 - ECONOMIC IMPERATIVE


# The economic imperative of the AI-native connected store.


Retailers are not short on technology. They are short on connected execution. Customer engagement, associate workflows, loss-prevention signals, store infrastructure, security telemetry, and AI pilots often operate in parallel rather than as one coordinated operating model.

Cisco Retail 2.0 reframes the business case around Total Economic Value: the value created when stores become connected enough for AI, automation, analytics, security, and human teams to act together. The model is benchmark-informed, customer-input-driven, and designed to support a disciplined executive value conversation.


## Total Economic Value is an envelope, not an ROI claim.



|What this figure is - and is not A TEV estimate combines modeled direct annual economic impact with strategic asset-value proxies and protected-value views. It is not a single-year profit number, EBITDA, net income, payback, TCO, or Cisco program ROI. Cost-side analysis belongs in a separate value-engineering engagement.|
|---|



|Layer A Modeled Direct Annual Economic Impact Annual economic impact from service efficiency, labor capacity, shrink reduction, energy savings, support efficiency, incident reduction, and other measurable drivers. Additive only after overlap controls.|Layer B First-Party Context Engine Strategic asset-value proxy for enriched profiles, consented messaging identities, known store sessions, product-intent signals, and retail media audience readiness.|
|---|---|
|Layer C Protected Value at Risk Downtime, cyber, safety, continuity, and trust exposure that the connected store foundation helps protect. Shown separately and non-additive by default.|Layer D Future Option Value Future capability value enabled by the connected-store foundation: AI, edge, computer vision, signage, workflow automation, and faster use-case rollout with less rework.|



|Layer B language control Layer B is a strategic asset-value proxy. It is not annual revenue, booked cash flow, ROI, or profit. It may become direct value only through activated and measured programs such as personalization, messaging, loyalty activation, retail media, product-intent follow-up, or supplier-funded campaigns.|
|---|


SECTION 02 - EXECUTIVE SUMMARY


# The connected retail operating model, in four value views.


For the CIO, CFO, COO, CISO, and retail leadership team, the connected store is not a technology refresh. It is a business operating model that changes how the store captures context, protects margin, supports frontline work, and underwrites digital resilience.


|What to hold onto|Why it matters|
|---|---|
|1. Direct impact is measurable, but must be net of overlap.|Layer A captures annual operating impact. The model separates gross impact from Net Layer A after de-duplication so the same customer journey, labor pool, incident, or asset base is not counted twice.|
|2. First-party context is a strategic asset.|Layer B values the data estate created when stores convert anonymous physical behavior into known, consented, CRM-ready customer and product context. It stays separate from annual revenue.|
|3. Resilience underwrites the value story.|Layer C does not simply add another number. It protects the store workflows, customer journeys, identity layers, and digital systems that make Layers A and B possible.|
|4. AI-readiness creates future option value.|Layer D recognizes that reusable infrastructure can reduce future rework and accelerate new AI use cases, but it remains qualitative unless activated roadmap or avoided-cost data exists.|


SECTION 03 - THE THREE PRIORITIES


# Three priorities. One connected retail operating model.


The TEV model follows the same three priorities as the Cisco Retail 2.0 whitepaper. The goal is not to force every retailer into the same value mix. Grocery, apparel, pharmacy, QSR, electronics, convenience, and department-store formats will weight the value pools differently.


|Priority 01 CX Growth and Loyalty Convert store, app, messaging, loyalty, service, product data, and retail media into more continuous customer journeys. Core value pools: known-customer engagement, service deflection, retail media, product intent, and first-party context.|Priority 02 EX, Operations, and Margin Protection Reduce frontline friction, protect associate capacity, improve service flow, automate repetitive work, strengthen safety, and elevate loss prevention into operational visibility.|Priority 03 IT Resilience, Trust, and Protection Build the secure, observable, AI-ready foundation needed to scale connected retail without increasing fragility. Core value pools: incident reduction, faster remediation, truck-roll reduction, cyber resilience, and protected continuity.|
|---|---|---|


SECTION 04 - HERO BENCHMARKS


# The benchmark library, mapped to the right value domains.


These proof points are directional anchors for scenario modeling. They are not universal Cisco-attributable results. Customer data, deployment scope, measurement maturity, and overlap controls determine the scenario output.


|40% Higher in-store conversion potential When shoppers use digital tools during the store journey; used as CX context, not a default uplift.|$1.84 vs $13.50 Self-service vs assisted support Directional service-cost gap for routine inquiry automation and cost-per-resolution planning.|1.6% / $112.1B Retail shrink baseline NRF/LPRC 2023 shrink context for margin protection and connected LP discovery.|$796B Total retail loss context Appriss 2026 Total Retail Loss benchmark context across shrink, returns, and broader loss exposure.|
|---|---|---|---|
|70-80% Routine inquiry automation potential Directional AI-assist planning range; requires containment, escalation, and customer service mix validation.|42% Task completion improvement signal Directional connected automation proof point; use with customer task and labor baselines.|40-66% Lighting / energy savings range Smart controls and centralized energy management proof-point range; apply only to lighting or verified energy scope.|43% Unplanned downtime reduction signal Observability benchmark context; use as protected-value context unless customer provides downtime exposure.|
|85% MTTR improvement signal Full-stack observability proof point for faster detection and resolution, not a guarantee of outage elimination.|80% Account takeover risk reduction signal Security proof point for identity, detection, and threat protection; treat as protected value context.|2.3x AI-leader growth signal Directional macro evidence that AI leaders outperform; not an annualized P&L assumption.|505% Vendor-commissioned CDP TEI ROI Useful proof that unified profiles can create value, but not a universal in-store personalization result.|


SECTION 05 - PRIORITY 01 CUSTOMER EXPERIENCE


# CX driving growth and loyalty.


The physical store must operate with more of the agility, relevance, and measurability of digital channels. Cisco Retail 2.0 frames CX value around the First-Party Context Engine: the ability to convert store traffic, app usage, Wi-Fi opt-in, messaging engagement, product scans, service interactions, and retail media exposure into governed customer context.


|Priority 1 value thesis CX value is created when the store stops being a blind spot. Known sessions, product intent, permissioned messaging, app engagement, and store behavior become inputs for better service, better offers, retail media, journey continuity, and richer first-party data.|
|---|



|Value driver|What it does|Hero signal|
|---|---|---|
|Personalized In-Store Engagement|Connect store presence, loyalty, app, Wi-Fi, and customer context to make the physical journey more relevant.|40% higher in-store conversion potential on contextualized interactions.|
|Omnichannel Journey Orchestration|Coordinate store, app, pickup, service, messaging, loyalty, and contact center interactions.|1.5x spend multiplier for omnichannel shoppers vs. single-channel peers.|
|AI-Powered Service and Outreach|Automate routine service moments and escalate complex needs with context.|$1.84 vs $13.50 self-service vs assisted support planning benchmark.|
|Dynamic Digital Merchandising and Retail Media|Use signage, ESLs, media surfaces, inventory, traffic, and content to make the store more adaptive and measurable.|0.5-2.5% revenue uplift potential from dynamic pricing and promotion context; in-store retail media standards support measurable activation.|
|Shelf-Edge Scan-to-Learn|Use SKU-level 2D codes to launch product education, offers, AI Q&A, or service flows.|GS1 2D / Digital Link migration gives retailers a path to turn shelf-edge engagement into measurable product-intent signals.|


Narrative: Priority 1 is not only about personalization. It is about creating a customer context layer that can serve multiple economic pathways: conversion influence, service deflection, loyalty recapture, retail media audiences, product education, and post-visit engagement. Cisco contributes the connective foundation: secure Wi-Fi, location intelligence, Webex engagement, app and API assurance, observability, security, and telemetry correlation.

SECTION 06 - PRIORITY 02 EMPLOYEE EXPERIENCE AND OPERATIONS


# EX driving workforce, operations, and margin protection.


Retail operations are under pressure from labor scarcity, margin pressure, shrink, return fraud, self-checkout complexity, energy cost, safety concerns, and growing device density. The connected store improves EX by reducing the operational friction that associates and store leaders absorb every day.


|Priority 2 value thesis The highest-leverage EX moves are often invisible to customers: fewer repetitive questions, faster task routing, less manual shelf work, better camera/POS context, fewer failed devices, and stronger store visibility. The result is a calmer, more controllable operating model.|
|---|



|Value driver|What it does|Hero signal|
|---|---|---|
|Margin Protection and Connected Loss Prevention|Connect cameras, POS events, safety workflows, self-checkout exceptions, and analytics to reduce loss exposure and improve operational visibility.|1.6% shrink / $112.1B U.S. retail shrink; $796B total retail loss context.|
|Intelligent Labor and Service Flow|Use traffic, queue, task, and fulfillment signals to improve store execution and service response.|20% throughput improvement and up to 70% wait reduction directional benchmark.|
|AI-Assisted Workflow Protection|Deflect routine questions and protect associate capacity.|70-80% routine inquiries auto-resolved in AI-assist planning context.|
|Task Automation and ESLs|Reduce manual price, promotion, and shelf-execution work.|42% task completion improvement signal from connected store automation.|
|Connected Facilities and Smart Lighting|Use lighting, sensors, PoE, occupancy, and facilities telemetry to improve energy, safety, sustainability, and operations.|40-66% lighting or energy reduction potential depending on scope and controls.|
|Connected Device Reliability|Improve uptime and performance for handhelds, POS, scanners, cameras, signage, ESLs, kiosks, and associate tools.|Truck-roll reduction and MTTR improvement become direct operating levers when device incidents are visible remotely.|


Narrative: Priority 2 is where the connected store protects the operating line. Shrink, return fraud, self-checkout leakage, labor interruptions, task burden, energy use, and device failures all erode execution. Cisco helps connect the sensing, networking, video, workflow, security, and observability layers that allow stores to see friction earlier and respond with more consistency.

SECTION 07 - PRIORITY 03 IT FOUNDATION


# IT foundation of resilience, trust, and protection.


Priority 3 underwrites everything above it. Customer engagement, associate productivity, loss prevention, retail media, signage, ESLs, smart lighting, cameras, Webex workflows, and AI-enabled use cases all depend on secure, observable, resilient infrastructure.


|Priority 3 value thesis IT is not an adjacent line item. It is the substrate that determines whether CX and EX can scale without increasing fragility. The same foundation that supports POS must also support cameras, sensors, signage, associate devices, cloud applications, security, observability, and AI workloads.|
|---|



|Value driver|What it does|Hero signal|
|---|---|---|
|Secure Connected Store Foundation|Provide reliable connectivity for stores, devices, applications, cloud services, and critical workflows.|Cisco-commissioned Meraki TEI and related sources support cloud-managed operations, remote visibility, and faster troubleshooting.|
|Zero Trust and Store Segmentation|Protect users, devices, applications, IoT, guest Wi-Fi, vendors, and payment-related systems.|Breach and credential-risk benchmarks support treating segmentation as protected value, not new revenue.|
|Observability as a Business Capability|Connect technical performance to customer, associate, and store workflow impact.|85% MTTR and 43% unplanned downtime reduction signals from observability proof points.|
|Cyber Resilience and Risk Reduction|Reduce exposure to ransomware, account takeover, credential theft, vendor risk, and shadow AI.|80% account-takeover risk reduction signal and broader cyber cost benchmarks support protected-value modeling.|
|AI-Ready Edge and Cloud Infrastructure|Support distributed AI workloads, cameras, sensors, APIs, edge workflows, and real-time store intelligence.|2.3x faster sales growth for AI leaders as directional macro context, not a direct P&L assumption.|


Narrative: Priority 3 creates direct IT efficiency through fewer incidents, faster diagnosis, lower support effort, fewer truck rolls, and reduced remediation burden. It also creates non-additive protected value by reducing exposure to outage, cyber, data, workflow, and operational continuity risk. In the TEV model, this layer protects the value story rather than inflating the headline number.

SECTION 08 - PROTECTED VALUE AT RISK


# The non-additive layer that underwrites the entire envelope.


Protected value should not be collapsed into a single additive TEV number unless the exposure pools are clearly separated. For most early-stage scenarios, Layer C is presented as a set of protected-value views: what could be disrupted, what Cisco helps protect, and what customer data would be required to quantify the exposure.


|Protected surface|Protection mechanism|
|---|---|
|Omnichannel continuity|Protects BOPIS, curbside, delivery, returns, messaging, and service workflows from API, cloud, store-network, and SaaS dependency failures.|
|Store operations and device reliability|Protects POS, handhelds, scanners, cameras, signage, ESLs, kiosks, associate devices, and operational systems from performance degradation and outages.|
|Cyber resilience and trust|Protects customer data, loyalty accounts, payment-adjacent systems, vendor access, cloud applications, identities, and connected devices.|
|Safety and workplace continuity|Supports safety visibility and escalation workflows. Quantification requires customer incident, workers compensation, turnover, claims, legal, insurance, or other cost data.|
|AI and data governance|Protects the data flows, APIs, identity layers, telemetry, and governance controls needed for AI-enabled customer, associate, and operational workflows.|



|Non-additive rule Layer C underwrites the value envelope. It is not added to Layer A unless the customer provides distinct, non-overlapping exposure pools and the scenario explicitly explains the protected-value method.|
|---|


SECTION 09 - METHODOLOGY AND CFO APPENDIX


# Finance-grade logic, executive-ready hygiene.


The TEV model is designed to start a credible value conversation, not replace a value-engineering engagement. It separates value layers, applies deployment scope, distinguishes hard savings from capacity and protected value, and requires overlap controls before showing Net Layer A.


|Method component|How the page should explain it|
|---|---|
|A. Direct Annual Economic Impact|Layer A includes measurable annual economic value from service efficiency, labor capacity, shrink reduction, energy savings, support efficiency, incident reduction, and other drivers. Output should show Gross Layer A before overlap, overlap adjustment, and Net Layer A after overlap.|
|B. First-Party Context Engine|Layer B is a strategic asset-value proxy for known sessions, enriched profiles, consented messaging identities, product intent, and retail media audience readiness. It is range-based and always separated from P&L claims.|
|C. Protected Value at Risk|Layer C is shown separately as continuity, cyber, safety, downtime, and trust protection. It should generally not be added to the TEV envelope.|
|D. Future Option Value|Layer D captures the option value of reusable AI-ready infrastructure. It remains qualitative unless activated roadmap, avoided rework, or use-case-specific cost data exists.|



## Key caveats the page should preserve.


- TEV is an envelope, not a P&L claim. Do not describe the output as single-year profit, EBITDA, net income, payback, or ROI.

- Net means net of overlap, not net income. Net Layer A means modeled impact after de-duplication / overlap controls.

- Revenue uplift is not automatically profit. If a customer wants P&L translation, that belongs in a separate finance conversation with margin assumptions.

- Vendor-commissioned studies must be labeled. TEI and case-study sources can support proof points, but should not be presented as universal benchmarks.

- Optional modules require inputs. SCO, return fraud, OSA, safety, downtime, cyber, and AI-readiness quantification require customer-specific baselines or should remain directional.

SECTION 10 - SCENARIO INPUTS


# What inputs improve the estimate?


The model can run a first-pass directional scenario with partial data, but confidence improves quickly when the retailer provides operating baselines. Unknown values should be handled with conservative defaults and clearly labeled assumptions.


|Input category|Examples|
|---|---|
|Retail scale|Annual revenue, store count, store formats, deployment scope, average basket, annual visits or transactions.|
|Customer and identity|Known customer rate, active loyalty/app users, messaging-reachable customers, Wi-Fi opt-in, consented identities, app engagement.|
|CX and service|Annual service contacts, routine inquiry percentage, cost per assisted contact, cost per resolution, pickup/return/service recovery volume.|
|Retail media|Retail media maturity, addressable impressions, screen/shelf inventory, supplier demand, fill rate, attribution maturity.|
|Loss prevention|Shrink rate, return value, self-checkout sales mix, return fraud/abuse baseline, LP investigations, investigation hours, LP labor rate.|
|Operations and facilities|Price-change cadence, tag counts, task hours, utility spend, lighting share, operating hours, facilities strategy.|
|IT and resilience|Store-impacting incidents, MTTR, support labor rate, truck rolls, cost per truck roll, downtime exposure, cyber exposure, sensitive systems.|
|AI roadmap|Activated AI use cases, deployment roadmap, avoided rework, edge/cloud requirements, data governance maturity.|


SECTION 11 - EXAMPLE OUTPUT STRUCTURE


# What the TEV output should show.



|Layer|Output treatment|
|---|---|
|Layer A|Net modeled annual operating impact range after overlap controls. Show gross, overlap adjustment, and net where useful.|
|Layer B|Strategic First-Party Context Engine asset-value proxy range, with the Layer B disclaimer next to the number.|
|Layer C|Protected value at risk, quantified only with sufficient customer exposure data; otherwise shown as protected-value views.|
|Layer D|Future option value, usually qualitative unless roadmap, avoided rework, or activated use-case assumptions are provided.|



|Do not add the layers automatically Layer A, Layer B, Layer C, and Layer D are different value types. The page may present them together as an economic opportunity surface, but it should not imply that all layers are directly additive P&L value.|
|---|


SECTION 12 - NEXT STEP


# From value hypothesis to customer scenario.


The TEV page should help Cisco teams, partners, and retail stakeholders move from strategic interest to a structured first-pass scenario. The next step is not to ask the customer to buy the entire stack. It is to identify which value pools matter most, collect the inputs that improve confidence, and determine where Cisco teams, partners, value analysts, and customer stakeholders should go deeper.

- Start with the retailer’s top priority. CX growth, operations and margin protection, or IT resilience.

- Collect the minimum scenario inputs. Revenue, store count, deployment scope, identity/customer data, service volume, shrink/loss data, utility spend, and IT incident baselines.

- Run a conservative/base/upside TEV scenario. Keep Layer A, Layer B, Layer C, and Layer D separated.

- Use the output to guide discovery. Prioritize customer data gaps, high-confidence value drivers, and areas requiring value analyst follow-up.


|Suggested web CTA Start a first-pass TEV scenario using customer-provided inputs. Use the result to identify which connected-store value pools deserve deeper discovery with Cisco and the customer value team.|
|---|


APPENDIX - SOURCE NOTES


# Source notes for page buildout.


The web implementation should link detailed benchmark and methodology references through the companion TEV source registry. The page copy can use short source labels rather than dense footnotes in every section.


|Source area|Recommended treatment|
|---|---|
|Retail shrink and LP|NRF / LPRC 2023 National Retail Security Survey; Appriss Retail 2026 Total Retail Loss Benchmark; Appriss / Deloitte returns research; ECR Retail Loss Group self-checkout research.|
|Service and contact economics|Zendesk / Forrester TEI; secondary self-service vs assisted-support cost benchmark; customer cost-per-resolution data should replace benchmarks when available.|
|Retail media and digital merchandising|IAB in-store retail media standards; retailer-led networks and in-store media market activity; use Grocery TV only as market proof unless a specific Cisco integration is confirmed.|
|Cisco and vendor-commissioned proof points|Cisco Meraki TEI, Cisco Full-Stack Observability TEI, ThousandEyes, AppDynamics, Splunk, Amperity TEI, Twilio Segment TEI, Charles TEI. Label commissioned studies clearly.|
|Cyber and resilience|IBM Cost of a Data Breach, Verizon DBIR, Sophos retail ransomware research, Cisco security sources. Treat primarily as protected-value context.|
|AI macro context|McKinsey, Deloitte, Capgemini, BCG, Walmart, Cisco, ThousandEyes, and other AI-native retail research should support the thesis but not become direct formula defaults.|


Prepared as page copy for the Cisco Retail 2.0 TEV landing page webapp. This document is not the calculator engine and does not include full formulas, source IDs, or scenario math.

---

# 6. TEV Calculator v1.11

# Cisco Retail 2.0 TEV Calculator

## Retail in the AI-Native Era: Connected Store Operating Model

**Document type:** Portable conversational-agent calculator and TEV study guide  
**Version:** v1.11  
**Purpose:** Enable a conversational agent to estimate the Total Economic Value (TEV) of a Cisco-enabled connected store operating model for an enterprise retailer.

**v1.11 change note:** Small calibration patch based on the v1.10 multi-archetype test. Updates the return-fraud materiality gate so benchmark-derived return-fraud values are suppressed when they exceed 10% of Net Layer A or become a top-3 Layer A driver, and adds a stricter QSR / fast casual / convenience known-session dampening override for high-frequency transactional formats. Core TEV architecture, source registry, v1.10 segment controls, and all source-governance rules are preserved.

**v1.10 change note:** Segment-mechanics hardening patch based on the v1.9 multi-archetype test. Adds a hard return-fraud gating rule when benchmark-derived return value becomes material, explicit A1 basket-tier guidance, numeric B1 known-session frequency dampening factors, QSR-specific investigation-hour guidance, loyalty-active-rate clarification, explicit A4/B6 retail-media maturity gating, a shared truck-roll allocation rule across A11/A12/A13, and segment-aware C4 safety discovery prompts. Core TEV architecture, source registry, and v1.9 segment controls are preserved.

**v1.9 change note:** Segment-guidance patch based on the multi-archetype AI-bot test. Adds segment-aware return-rate defaults, QSR-specific shrink/waste guidance, retail-media maturity gating, investigation-hour tiering for high-volume LP teams, Layer B known-session frequency dampening, `loyalty_active_rate`, deployment-order clarification, A6 internal-overlap cap clarification, and output requirements for any driver contributing more than 10% of Net Layer A. Core TEV layer structure, source registry, and v1.8.1 output-consistency controls are preserved.

**v1.8.1 change note:** Output-consistency patch based on the first AI-bot test run. Adds explicit rules requiring the executive layer summary to show **Net Layer A after overlap**, top-driver values to match detailed driver sections, every included driver to appear in a driver-level breakdown, Layer B headlines to use the selected scenario range rather than blended ranges, A6 umbrella caps to use modeled loss pools rather than raw revenue throughput, and confidence scoring to distinguish overall confidence from driver-level confidence. Core TEV math, sources, and model controls are unchanged.

**v1.8 change note:** Source-registry hygiene and citation-validation pass based on the May 2026 source audit. Resolves A3/A8 service-source gaps by adding official Zendesk TEI and a traceable self-service-versus-assisted-contact benchmark source, promotes Appriss 2026 TRL from manual-verification status to verified planning-range support, adds cost-per-resolution guidance to prevent contact-deflection overstatement, adds optional inventory/RFID/OSA source support without expanding default formulas, and reframes downtime benchmarks as Layer C context rather than hard defaults. Core TEV math and v1.7 model controls are unchanged.

**v1.7 change note:** Field-readiness polish and source-verification discipline pass. Adds Layer B escalation language for very large strategic-proxy values, strengthens C4 workplace-safety input gating, clarifies multi-pair overlap aggregation, removes unresolved OPS sources from preferred source-to-impact citations, adds manual-verification visibility guidance, and adds NRF 2023 data-currency context.

**v1.6 change note:** Model-control and Loss Prevention hardening pass. Tightens the A6 Margin Protection / Connected Loss Prevention module with an LP umbrella cap, explicit LP submodule overlap rules, revised conservative/base/upside planning ranges, an explicit self-checkout basis-point example, default-OFF treatment for SCO and phantom inventory unless customer data is provided, and movement of safety quantification fully into Layer C. Adds Layer B range/disclaimer output rules and strengthens finance-sensitivity guardrails. Core TEV layer structure and Layer B asset-value philosophy are unchanged.

---

# 1. How to use this file

A user should upload this Markdown file to a conversational agent and provide retailer-specific inputs such as annual revenue, store count, traffic, average basket, contact volume, shrink rate, store labor assumptions, energy spend, IT incident volume, and customer identity maturity.

The agent should then produce a structured TEV estimate across four layers:

| Layer | Name | Treatment |
|---|---|---|
| **Layer A** | Modeled Direct Annual Economic Impact | Current-period economic value from measurable drivers. Additive only after overlap controls. |
| **Layer B** | First-Party Context Engine Strategic Value Proxy | Strategic proxy for known customers, CRM-ready identities, product intent, messaging relationships, and addressable media audiences. Non-booked revenue unless activated and measured. |
| **Layer C** | Protected Value at Risk | Resilience, cyber, uptime, workflow-continuity, and risk-reduction value. Shown separately and non-additive by default. |
| **Layer D** | Future Option Value | Future capability value enabled by the connected-store foundation. Usually qualitative unless activated, contracted, resourced, measured, and governed. |

This is a **Total Economic Value model**, not an ROI, TCO, gross-margin, net-margin, or profit model.

**Scope note:** This file is a portable TEV and connected-store knowledge base. It is intended to help Cisco teams, partners, and retail leaders explore directional value pools quickly in a conversational agent. It is not a final value-engineering business case, implementation plan, TCO, NPV, payback model, or procurement-ready ROI analysis.

**Steady-state disclosure:** Unless the user explicitly asks for rollout timing, figures should be presented as steady-state annual TEV at the specified deployment scope. They do not reflect Year 1 ramp, implementation timing, program cost, or phased rollout economics. Those topics belong in a follow-on value analyst / finance engagement.

---

# 2. Core modeling rules

## 2.1 TEV is not ROI or TCO

Do not calculate ROI, payback, profit contribution, gross margin, contribution margin, EBITDA, or net operating profit by default.

The core TEV output should remain:

1. **Layer A:** Modeled direct annual economic impact  
2. **Layer B:** First-Party Context Engine strategic value proxy  
3. **Layer C:** Protected value at risk  
4. **Layer D:** Future option value  

If the user asks a finance-specific follow-up such as “What portion of this could show up as profit?” or “What would this look like as ROI?”, treat that as a separate conversation. See Section 16: Profit / ROI / TCO Conversation Guardrail.

## 2.2 Do not collapse the layers into guaranteed ROI

Do not add Layers A, B, C, and D together and call the result guaranteed ROI, booked revenue, profit, or savings.

Use language such as:

- “TEV envelope”
- “modeled direct annual economic impact”
- “strategic value proxy”
- “protected value at risk”
- “future option value potential”
- “benchmark-informed estimate”
- “customer-specific assumptions required”

Avoid language such as:

- “guaranteed ROI”
- “guaranteed returns”
- “profit” unless the user explicitly requests profit sensitivity
- “all value layers are additive”
- “protected value is new revenue”
- “Layer B is booked revenue”
- “Layer D is current-year value”

## 2.3 Use value-driver modeling, not use-case-by-use-case modeling

Model the TEV at the **value-driver level**. Use cases should sit underneath value drivers as examples.

Structure:

```text
Priority -> Value Driver -> Use Cases -> Impact Metric(s) -> Formula -> Output
```

Do not create separate formulas for every use case unless the user provides specific data that isolates that use case.

## 2.4 Use customer inputs first, defaults second

If the customer provides data, use it. If not, use conservative benchmark defaults and clearly label them as defaults.

Always state:

- which inputs were customer-provided;
- which inputs were estimated;
- which benchmark defaults were used;
- which assumptions most affect the result;
- what additional information would improve confidence.

## 2.5 Apply overlap controls

Connected-store value pools overlap. Apply haircuts before presenting net Layer A.

Common overlap risks:

| Overlap area | Risk |
|---|---|
| Personalization + retail media + shelf-edge scans | Same transaction may be influenced by multiple CX drivers. |
| AI service deflection + workflow protection | Same routine inquiry pool may be counted twice. |
| Queue improvement + labor productivity | Same associate time or transaction recovery may be counted twice. |
| Shrink reduction + inventory accuracy | Same loss pool may be counted twice. |
| Observability + secure foundation + device reliability | Same incident pool may be counted across multiple IT drivers. |
| Zero Trust + cyber resilience | Same breach exposure may be counted twice. |

---

# 3. Minimum input collection

Start with a short input request. Do not overwhelm the user.

Ask first:

1. Annual revenue
2. Number of stores
3. Annual store visits or transaction count, if known
4. Average basket / transaction value
5. Shrink rate, if known
6. Annual customer contact volume, if known
7. Store associate fully loaded hourly cost, if known
8. Annual utility spend, if known
9. Annual IT incident / outage volume, if known
10. Customer identity maturity: low / medium / high
11. Deployment scope: all stores or percent of stores in scope

If the user only provides annual revenue and store count, create a directional TEV estimate using conservative defaults and label it clearly.

---

# 4. Input dictionary

## 4.1 Company scale inputs

| Input | Description | If unknown |
|---|---|---|
| `annual_revenue` | Total annual retail revenue | Required if possible |
| `store_count` | Number of physical stores | Required if possible |
| `annual_store_visits` | Total annual physical visits | Estimate from transactions if needed |
| `annual_transactions` | Total annual store transactions | Estimate from revenue / average basket if needed |
| `average_basket` | Average in-store transaction value | Use segment default or ask user |
| `deployment_pct` | % of stores / estate in scope | Ask first. Use 100% only if the user explicitly says all stores, enterprise-wide, or full estate. If scope is not specified, label the output as full-estate potential, not a deployment plan. |

## 4.2 Customer identity and engagement inputs

| Input | Description | Default guidance |
|---|---|---|
| `loyalty_members` | Total loyalty members | Optional |
| `loyalty_active_rate` | % of loyalty members active / reachable during the modeled period | Use customer data. If unknown, conservative 30%, base 50%, upside 70%. If the user provides active loyalty / app users directly, use that number and do not apply this factor again. |
| `active_app_users` | Active app users | Optional |
| `email_database_size` | CRM / email reachable customers | Optional |
| `messaging_identities` | SMS / RCS / WhatsApp / phone-based reachable identities | Optional |
| `identified_session_rate` | % of store visits tied to loyalty, app, Wi-Fi, phone, email, or CRM | Use 5% conservative, 15% base, 30% upside if unknown |
| `wifi_onboarding_rate` | % of store visitors who connect / identify through Wi-Fi | Optional |
| `scan_engagement_rate` | % of visits or product encounters generating QR / GS1 2D / shelf-edge scans | Use conservative assumptions unless customer has data |
| `opt_in_match_rate` | % of scans or Wi-Fi sessions that become CRM-ready | Use 10% conservative, 25% base, 40% upside if unknown |
| `qualified_intent_rate` | % of scans/Q&A that produce useful product intent | Use 25% conservative, 50% base, 70% upside if unknown |
| `retail_media_maturity` | Retail media / in-store media program maturity: none, emerging, limited, active, or mature | Used to gate A4 Layer A retail media value and B6 audience value. See Section 4.8 and A4. |

## 4.3 Contact center and service inputs

| Input | Description | Default guidance |
|---|---|---|
| `annual_contacts` | Total annual customer contacts | Estimate only if needed |
| `routine_inquiry_pct` | % addressable by AI / automation | Conservative 20%, base 35%, upside 50% |
| `cost_per_assisted_contact` | Cost per phone/chat/email/live support contact | Use customer data; if unavailable, benchmark cautiously |
| `deflection_pct` | % of routine contacts deflected or automated | Conservative 10-15%, base 15-25%, upside 25-40% |
| `aht_minutes` | Average handle time | Optional |
| `agent_cost_per_minute` | Agent labor cost per minute | Optional |

## 4.4 Store labor and operations inputs

| Input | Description | Default guidance |
|---|---|---|
| `associate_hourly_cost` | Fully loaded hourly cost | Use customer data if possible |
| `weekly_price_change_hours_per_store` | Manual price-change / tag-change hours per store per week | Conservative 4-6, base 8-12, upside 12-20+ |
| `routine_store_inquiries_per_week` | Routine customer/associate inquiries per store per week | Optional |
| `minutes_saved_per_inquiry` | Minutes saved per deflected inquiry | Conservative 1-2, base 2-4, upside 4-6 |
| `queue_abandonment_rate` | % of customers at risk of walking out due to queues | Optional |
| `bopis_pickup_volume` | BOPIS / pickup order volume | Optional |
| `returns_volume` | Returns volume | Optional |
| `device_incidents` | Annual device-related incidents | Optional |
| `cost_per_support_ticket` | IT / store support ticket cost | Optional |
| `truck_roll_count` | Annual store IT dispatches | Optional |
| `cost_per_truck_roll` | Field technician visit cost | Optional |

## 4.5 Shrink, loss prevention, returns, self-checkout, inventory, and safety inputs

| Input | Description | Default guidance |
|---|---|---|
| `shrink_rate` | Shrink as % of revenue or inventory value. Prefer sales-based shrink if using the default formula. | Use customer data if possible. If unknown, use 1.10% conservative, 1.60% base, 2.10% high-risk/upside. |
| `addressable_shrink_pct` | % of total shrink that is addressable by connected LP, AI video, POS correlation, process controls, investigation workflow, and operational controls. | Conservative 30%, base 50%, upside 65%. Treat 73% as a ceiling that requires source/customer justification. |
| `shrink_reduction_pct` | Reduction of addressable shrink. | Conservative 5-10%, base 10-15%, upside 15-20%. Cap at 20% in early-stage scenarios unless customer-confirmed program data supports more. |
| `annual_return_value` | Gross dollar value of returned merchandise at retail price. | Use customer-provided value. If unknown and returns are in scope, estimate as `annual_revenue × return_rate` and label as a default. |
| `return_rate` | Returned merchandise as % of revenue. | Use customer data. If unknown, use the segment-aware fallback table in Section 4.8 and label the result as benchmark-derived. |
| `return_fraud_abuse_rate` | % of return value estimated to be fraudulent or abusive. | Conservative 8.0%, base 13.2%, upside 15.14%. Use with caveat and source support. |
| `return_fraud_reduction_pct` | Reduction of addressable return fraud / abuse. | Conservative 5-10%, base 10-15%, upside 15-20%. Apply mandatory overlap vs. shrink unless populations are distinct. |
| `sco_sales_mix` | % of revenue or transactions processed through fixed SCO, Scan & Go, or mobile checkout. | Default OFF unless customer provides SCO mix or explicitly asks for SCO analysis. |
| `sco_loss_bps` | Loss basis points applied to revenue in SCO scope. | Conservative 1.0 bps, base 2.6 bps, upside 5.0 bps in early-stage scenarios. Higher values require customer data. |
| `annual_lp_investigations` | Annual LP investigation count. | Optional; ask if investigation labor is being modeled. |
| `baseline_hours_per_investigation` | Current average investigation time. | Use customer data. If unknown, use the investigation-volume tiering in Section 4.8. Do not default high-volume LP teams to 9 hours without caveat. |
| `investigation_time_saved_pct` | % reduction in investigation time from AI video search, POS-video correlation, or case automation. | Conservative 30%, base 50%, upside 70%. Higher values should be case-study support only. |
| `lp_labor_rate` | Fully loaded LP labor rate. | Use customer data; do not use upside investigation savings without customer labor input. |
| `osa_sales_in_scope` | Category, department, or omnichannel sales exposed to measured on-shelf availability / phantom inventory issues. | Default OFF unless customer provides OSA baseline, affected category sales, BOPIS failure, or stockout data. Do not default to total revenue. |
| `osa_improvement_pct` | Improvement in on-shelf availability or inventory accuracy for the affected scope. | Customer data preferred. Keep conservative if used. |
| `osa_sales_sensitivity_factor` | Sales sensitivity to OSA improvement. | Conservative 0.10%, base 0.30%, upside 0.60% for early-stage scenarios; default OFF without customer OSA data. |
| `safety_incident_exposure` | Worker safety, workplace violence, claims, insurance, turnover, or incident-cost exposure. | Layer C only unless customer provides specific cost data. |

**A6 input gating rules:**

- Do not enable the SCO submodule unless `sco_sales_mix` is provided or the user explicitly asks for a SCO scenario.
- Do not enable the phantom inventory / OSA submodule unless `osa_sales_in_scope` and an OSA / stockout / inventory-accuracy baseline are provided.
- Do not quantify safety in Layer A. Safety belongs in Layer C protected value unless customer-provided cost data supports a separate finance discussion.
- Use customer-provided loss pools first. Benchmark defaults are for directional scenario framing only.

## 4.6 Energy and facilities inputs

| Input | Description | Default guidance |
|---|---|---|
| `annual_utility_spend` | Total annual utility spend | Use customer data if possible |
| `lighting_share_pct` | % of utility spend attributable to lighting | Conservative 15%, base 25%, upside 35% if unknown |
| `lighting_savings_pct` | Savings on lighting-related energy | Conservative 10-15%, base 15-30%, upside 30-50%+ |
| `facilities_labor_savings` | Optional avoided facilities hours/cost | Optional |

## 4.7 IT, observability, and cyber inputs

| Input | Description | Default guidance |
|---|---|---|
| `annual_it_incidents` | Store-affecting IT/network/app/device incidents | Use customer data if possible |
| `current_mttr_hours` | Mean time to resolution | Optional |
| `mttr_reduction_pct` | Expected MTTR reduction | Conservative 10-20%, base 25-50%, upside 50-85% |
| `it_labor_cost_per_hour` | Fully loaded IT labor cost | Optional |
| `annual_downtime_hours` | Annual downtime hours or workflow disruption hours | Optional |
| `downtime_value_per_hour` | Value exposure per downtime hour | Optional |
| `cyber_exposure` | Estimated cyber/breach exposure | Optional |
| `incident_probability` | Annual probability factor | Optional |
| `risk_reduction_pct` | Risk reduction assumption | Optional |
| `observability_maturity` | Low / medium / high | Use to choose scenario |
| `zero_trust_maturity` | Low / medium / high | Use to choose scenario |


## 4.8 Segment guidance and default-discipline rules

Use customer-provided data first. Segment defaults are fallback assumptions for directional scenario framing only. They must be labeled as benchmark-informed defaults, not customer facts.

### 4.8.1 Segment-aware return-rate fallback

Use `annual_return_value` when available. If not available and the return-fraud module is in scope, estimate `annual_return_value = annual_revenue × return_rate` using a segment fallback, then disclose the assumption.

| Segment | Conservative fallback | Base fallback | Upside / high-return context | Notes |
|---|---:|---:|---:|---|
| Grocery / supermarket | 1% | 2% | 3% | Usually lower merchandise-return exposure; exclude spoilage/waste unless customer confirms returns treatment. |
| Pharmacy / drugstore | 3% | 5% | 8% | Use only for front-store merchandise unless the customer provides Rx-specific return / reversal data. |
| Fashion / apparel / footwear | 20% | 30% | 40% | High return exposure; strongly prefer customer-provided return value. |
| Department store / mass apparel | 15% | 22% | 30% | Blend of apparel and general merchandise; prefer customer data. |
| Electronics / big-box specialty | 8% | 10% | 12% | Return fraud can be material; customer return value is high priority. |
| QSR / restaurant / fast casual | 0% | 0.25% | 1% | Do not model normal merchandise returns unless customer has refunds/appeasements data. |
| Home / furniture / specialty hardlines | 5% | 8% | 15% | Return logistics may be costly; use customer category data if possible. |

**Return-rate guardrail:** Segment fallback return rates may enable a directional return-fraud calculation, but they should not be used for a proposal-quality business case. If return fraud becomes a top-5 Layer A driver, ask for customer-provided `annual_return_value` before showing externally.

**v1.11 hard gate:** If `annual_return_value` is benchmark-derived, suppress the modeled return-fraud value from the executive Net Layer A total when either condition is true:

1. the modeled return-fraud reduction exceeds **10% of Net Layer A**, or
2. return fraud would become a **top-3 Layer A value driver**.

When the gate triggers, show return fraud as a **requires customer confirmation** opportunity and list `annual_return_value`, actual return rate, and fraud / abuse baseline as required discovery inputs. The agent may show the directional calculation in an assumptions appendix, but it must not count it in the executive Net Layer A summary until customer-provided return data is available.

### 4.8.2 Segment-aware shrink / waste guidance

| Segment | Default treatment when customer shrink data is missing |
|---|---|
| Grocery / supermarket | Use 1.60% base or 2.10% high-risk where SCO, fresh loss, or high-theft categories are significant. Separate spoilage/waste from theft if the customer can provide detail. |
| Pharmacy / drugstore | High-risk segment; 2.10% may be appropriate as a benchmark-informed upper planning assumption. |
| Mass merchant / supercenter | High-risk segment; 1.60% base to 2.10% high-risk depending on format. |
| Fashion / apparel / department store | Start with 1.60% base unless customer indicates high-theft categories or fitting-room vulnerability. |
| Electronics / big-box specialty | 1.60% base to 2.10% high-risk; return fraud and warranty / claims issues may be as important as store shrink. |
| QSR / restaurant / fast casual | Do **not** use general retail shrink by default. Use 0.50% - 1.00% as a directional food waste / process-loss proxy only when the user wants an operations-loss scenario. Treat addressability differently: portion control, spoilage, inventory accuracy, and process compliance, not classic LP shrink. |

### 4.8.3 Investigation-hour tiering

The 9-hour investigation benchmark can overstate per-case labor for high-volume LP teams. Use customer-provided baseline hours first. If unknown, apply this tiering:

| Annual LP investigations | Default baseline hours per case | Use case |
|---:|---:|---|
| Fewer than 30,000 | 7-9 hours | Lower-volume, more complex investigations. |
| 30,000-80,000 | 5-7 hours | Mixed case volume; use midpoint unless customer data is available. |
| More than 80,000 | 3-5 hours | High-volume LP operations; assume triaged / shorter average cases unless customer confirms longer investigations. |

If `baseline_hours_per_investigation` is benchmark-derived and investigation savings becomes a top-5 Layer A driver, request customer validation before external use.

**QSR / fast-casual investigation note:** For QSR, fast casual, restaurant, and similar foodservice operators, do not automatically apply long-form LP investigation benchmarks. Cash-handling, refund, coupon, waste, and process-compliance investigations are often shorter than full video-based retail LP investigations. If the customer does not provide baseline hours, use **3-5 hours per case** for QSR / restaurant scenarios unless the customer confirms complex video-based investigations.

### 4.8.4 Retail media maturity gating

Use `retail_media_maturity` to gate A4 Layer A retail media calculations.

| Maturity | A4 Layer A treatment | B6 Layer B treatment |
|---|---|---|
| None / not applicable | A4 retail media OFF. Keep only non-monetized signage / merchandising if customer provides data. | B6 OFF by default; qualitative future-option discussion only unless the user provides addressable audience and activation plan. |
| Emerging | Conservative only. Do not use base/upside CPM or category-lift assumptions without active campaigns and screen inventory. | Conservative-only B6 range. Do not use base/upside audience proxy without known audience, touchpoints, and activation readiness. |
| Limited | Conservative to base. Require at least estimated impressions, fill rate, and supplier demand. | Conservative/base B6 allowed with caveat. Upside suppressed unless measured audience monetization exists. |
| Active | Base allowed. Upside requires test/control evidence, measured impressions, or campaign results. | Base B6 allowed; upside requires measured addressability / closed-loop activation. Avoid double-counting monetized media revenue and audience asset value. |
| Mature | Full scenario range may be shown if the customer has measured campaigns, closed-loop attribution, and supplier demand. | Full B6 scenario range may be shown, still separate from A4 monetized revenue and still subject to Layer B disclaimers. |

### 4.8.5 High-frequency known-session dampening for Layer B

For high-frequency segments such as QSR, grocery, convenience, and pharmacy, many known sessions may come from the same customer. To avoid over-weighting repeated visits, calculate or estimate `known_sessions_per_active_customer` when active customer count is available.

| Known sessions per active customer per year | Standard B1 treatment |
|---:|---|
| Fewer than 12 | Full selected-scenario B1 range may be used. |
| 12-30 | Apply a **0.85x** dampening factor to B1 session value or use the lower-middle of the selected range. |
| 30-60 | Apply a **0.70x** dampening factor to B1 session value or use the lower end of the selected range. |
| More than 60 | Apply a **0.50x** dampening factor and show B1 separately. Ask for customer data on unique visitors, loyalty active rate, and session value before using externally. |

**v1.11 QSR / convenience override:** For QSR, fast casual, convenience, fuel / c-store, and similar high-frequency transactional formats, apply stricter B1 dampening because many known sessions may represent repeated low-value visits by the same active customers. Use this override instead of the standard table for those formats:

| Known sessions per active customer per year | QSR / convenience B1 treatment |
|---:|---|
| Fewer than 12 | Apply a **0.85x** dampening factor unless customer evidence supports full base value. |
| 12-30 | Apply a **0.70x** dampening factor. |
| 30-60 | Apply a **0.50x** dampening factor. |
| More than 60 | Apply a **0.35x** dampening factor or keep B1 qualitative until customer-specific unique-customer, frequency, and activation-value data is provided. |

Do not use high-frequency B1 session value as a substitute for B2 enriched profile value; they are related but different value pools. If active customer count is not available, present B1 as a range and ask for unique customer / active loyalty data.

### 4.8.6 Basket-tier and frequency guidance for A1 personalization

A1 personalization uplift should not be treated as identical across low-basket, high-frequency formats and high-basket, considered-purchase formats. Use customer-measured activation data first. If unavailable, apply this guidance mechanically:

| Segment / basket pattern | A1 uplift treatment |
|---|---|
| Low basket / high frequency, including QSR, convenience, grocery, and some pharmacy traffic | Conservative to low-base only. Use the lower end of A1 uplift and apply B1 frequency dampening separately. |
| Mid-basket / moderate frequency retail | Base range allowed if identification, messaging, or app engagement is credible. |
| High-basket / considered purchase, including electronics, specialty hardlines, furniture, and complex product categories | Base range allowed; upside requires customer evidence such as measured recommendation lift, attach-rate lift, product-assist conversion, or campaign results. |
| Apparel / fashion with strong loyalty or styling engagement | Base range allowed; upside requires measured personalization, stylist, fit, or recommendation performance. |

If A1 becomes a top-5 Layer A driver and uplift is benchmark-derived, show the uplift percentage and ask for customer evidence before using the result externally.

### 4.8.7 Deployment-order rule

Unless a formula explicitly says otherwise, apply `deployment_pct` **last**, after eligibility, adoption, identification, match, addressability, and scenario factors. This avoids applying deployment scope twice or before the population has been properly filtered.

### 4.8.8 Segment testing requirement

When testing the calculator across multiple retailer types, the agent must report which segment-specific defaults were used and which drivers were turned OFF because the segment does not support them.

---

# 5. Scenario logic

Each driver should support three scenarios.

| Scenario | Treatment |
|---|---|
| **Conservative** | Low adoption, low uplift, high caution, larger overlap haircut |
| **Base** | Reasonable adoption, benchmark midpoint, standard overlap haircut |
| **Upside** | Strong adoption, strong integration, mature operating model, smaller but still explicit caution |

Default to **base** if the user asks for one estimate, but provide conservative/base/upside when possible.

**Upside discipline:** If more than three value drivers are set to upside assumptions, the agent must flag the scenario as aggressive and include this sentence: “This scenario uses upside assumptions across multiple drivers and should be treated as a high-end opportunity view, not the default planning case.”


---

# 6. Layer A: Modeled Direct Annual Economic Impact

Layer A includes measurable annual economic value from commercial actions, recovered revenue, avoided costs, workflow improvements, energy savings, shrink reduction, service deflection, and IT mitigation.

Layer A is the only layer that may be presented as direct annual modeled impact. It is still not guaranteed profit or ROI.

## 6.1 CX Value Driver A1: Personalized In-Store Engagement

**Purpose:** Estimate economic value from identified, context-aware store engagement.

**Use cases:**

- arrival-aware engagement
- loyalty/app store mode
- Wi-Fi-enabled known session recognition
- location-based offers
- app-to-store continuity
- known-customer follow-up
- contextual customer service triggers

**Formula:**

```text
Modeled CX Engagement Value =
Annual Store Visits
× Identified Session Rate
× Average Basket
× TEV Uplift %
× Deployment %
```

**Default TEV uplift range:**

| Scenario | Uplift |
|---|---:|
| Conservative | 0.25% - 0.75% |
| Base | 1.0% - 2.0% |
| Upside | 2.0% - 3.0%+ |

**Basket and frequency guidance:** Use the lower end of the uplift range for very low-basket, high-frequency segments such as QSR unless the customer has app, loyalty, or offer-response data. Use higher-end uplift only when the retailer has measured activation evidence, especially for high-basket categories such as electronics.

**Cisco relevance:** Cisco Spaces, Wi-Fi, Webex Connect, Webex AI, AppDynamics, ThousandEyes, Splunk.

**Double-counting warning:** Do not stack fully with retail media, shelf-edge scan conversion, or omnichannel lift if they influence the same transactions.

---

## 6.2 CX Value Driver A2: Omnichannel Journey Orchestration

**Purpose:** Estimate recovered economic value from better pickup, curbside, returns, substitution approval, and service recovery journeys.

**Use cases:**

- pickup readiness alerts
- curbside arrival coordination
- substitution approval flows
- proactive service recovery
- returns orchestration
- order-status automation
- contact center escalation with context

**Revenue / recovered journey formula:**

```text
Recovered Journey Economic Value =
At-Risk Omnichannel Revenue
× Recovery Rate
× Deployment %
```

**Service formula:**

```text
Service Savings =
Avoided Contacts × Cost per Contact
```

**Default recovery range:**

| Scenario | Recovery assumption |
|---|---:|
| Conservative | 0.10% - 0.30% of omnichannel revenue |
| Base | 0.30% - 0.75% |
| Upside | 0.75% - 1.50% |

**Cisco relevance:** Webex Connect, Webex AI, Webex Contact Center, AppDynamics, ThousandEyes, SD-WAN, Splunk.

**Double-counting warning:** Avoid stacking with personalization if both claim the same recovered sale. Avoid stacking contact savings with AI service deflection unless contact populations are separated.

---

## 6.3 CX Value Driver A3: AI-Powered Service and Outreach

**Purpose:** Estimate avoided cost from automated service, proactive outreach, AI-assisted responses, and contact deflection.

**Use cases:**

- routine inquiry automation
- product-originated conversations
- proactive pickup / return / delay messages
- AI-assisted customer service
- contact center escalation with context
- post-interaction summaries
- service recovery triggers

**Deflection formula:**

```text
Contact Deflection Savings =
Annual Contacts
× Routine Inquiry %
× Deflection %
× Cost per Assisted Contact
× Deployment %
```

**v1.8 cost-per-resolution guardrail:** When the user has data on repeat contacts, first-contact resolution, or escalation rates, prefer a cost-per-resolution view over a simple cost-per-contact view. Automation often deflects routine tier-1 contacts first, leaving more complex interactions for human agents. To avoid overstating savings, calculate avoided value on the successfully resolved routine inquiry population, not all contacts.

Optional refinement:

```text
Resolved Deflection Savings =
Annual Contacts
× Routine Inquiry %
× Deflection %
× Successful Resolution / Containment %
× (Assisted Cost per Resolution - Self-Service Cost per Resolution)
× Deployment %
```

If self-service and assisted-service cost benchmarks are used, cite `SRC-OPS-009` and disclose that the benchmark is a secondary source reporting a Gartner-style cost differential unless the customer provides their own cost-per-resolution data.

**AHT formula, if data is available:**

```text
AHT Savings =
Agent-Assisted Contacts
× AHT Reduction %
× Cost per Contact Minute
× Deployment %
```

**Default deflection range:**

| Scenario | Deflection |
|---|---:|
| Conservative | 10% - 15% |
| Base | 15% - 25% |
| Upside | 25% - 40% |

**Cisco relevance:** Webex Connect, Webex AI, Webex Contact Center, Splunk, AppDynamics, ThousandEyes.

**Double-counting warning:** Do not add full deflection, AHT reduction, and first-contact resolution improvement unless they apply to separate contact populations.

---

## 6.4 CX Value Driver A4: Dynamic Digital Merchandising and Retail Media

**Purpose:** Estimate economic value from in-store media, supplier-funded campaigns, digital signage, inventory-aware merchandising, and dynamic content.

**Layer decision rule:** Monetized or sponsor-funded retail media revenue belongs in Layer A when the retailer has an active or planned monetization program. Retail-media addressable audience value belongs in Layer B when valuing the strategic audience asset. Do not count both for the same audience unless the model clearly separates realized media revenue from additional unmonetized audience asset value.

**Retail media maturity gate:** Apply Section 4.8.4 before calculating A4. If maturity is `none`, A4 retail media is OFF. If `emerging`, use conservative assumptions only. If `limited`, use conservative/base only. If `active`, base is allowed and upside requires measured campaign evidence. If `mature`, the full scenario range may be shown with source and measurement caveats.

**Use cases:**

- in-store retail media
- digital signage monetization
- inventory-aware campaign rotation
- supplier-funded product education
- dynamic endcap / checkout / aisle content
- closed-loop campaign measurement
- store traffic-based merchandising

**Addressable impressions guidance:**

```text
Addressable Impressions =
Store Traffic in Scope
× Average Media Touchpoints per Visit
× Viewability / Presence Factor
× Fill Rate
× Deployment %
```

Use customer media platform data where available. If not available, label impression assumptions as directional.

**Media formula:**

```text
Retail Media Economic Value =
Addressable Impressions / 1,000
× CPM
× Fill Rate
× Media Value Factor
× Deployment %
```

**Category lift formula:**

```text
Modeled Merchandising Economic Value =
Category Sales in Scope
× Incremental Lift %
× Deployment %
```

**Default assumptions:**

| Variable | Conservative | Base | Upside |
|---|---:|---:|---:|
| In-store CPM | $10 | $25 | $50 |
| Media value factor | 50% | 60% | 70% |
| Category sales lift | 0.25% | 0.50% - 1.00% | 1.00%+ with test/control |

**Cisco relevance:** Cisco networking, PoE, Meraki, Cisco Spaces, Webex Connect, Splunk, AppDynamics, ThousandEyes.

**Double-counting warning:** Do not count both supplier-paid media revenue and gross product sales lift as the same economic value unless they are clearly distinct pools.

---

## 6.5 CX Value Driver A5: Shelf-Edge Scan-to-Learn

**Purpose:** Estimate selective Layer A value from scan-to-offer, scan-to-service, scan-to-education, or product-assist interactions. Most of this driver also feeds Layer B.

**Use cases:**

- QR / GS1 2D shelf-edge product education
- ingredient / allergen lookup
- compatibility checks
- recipes and solution-building
- coupon / offer actions
- product-originated AI Q&A
- request associate help
- escalation to Webex Contact Center

**Commercial action formula:**

```text
Scan-Originated Commercial Action Value =
Annual Scans
× Qualified Commercial Action Rate
× Average Action Value
× Deployment %
```

**Product Q&A containment formula:**

```text
Scan-to-Deflection Savings =
Product Q&A Interactions
× AI Containment Rate
× Avoided Contact Cost
× Deployment %
```

**Cisco relevance:** Webex Connect, Webex AI, Webex Contact Center, Cisco networking, AppDynamics, Splunk.

**Value-pool treatment:** A single scan can legitimately create multiple distinct value pools: (1) Layer B product-intent / declared-preference / CRM-ready identity value, (2) Layer A commercial action, upsell, coupon, basket, or service-deflection value, and (3) longer-term Layer B product/customer intelligence value. This is allowed when those pools are described separately.

**Double-counting warning:** Do not count the same signal twice inside Layer B. For example, do not count one allergen question as both a generic product-intent signal and a separate zero-party declared preference unless it represents two distinct captured data fields. Do not stack scan-driven commercial action value with personalization or retail media unless the transaction populations are clearly isolated.

---

## 6.6 EX Value Driver A6: Margin Protection and Connected Loss Prevention

**Purpose:** Estimate economic value from preventable / addressable loss pools, including shrink reduction, return fraud / abuse reduction, self-checkout exposure, investigation productivity, and optional inventory accuracy / on-shelf availability improvement.

**Important v1.6 control:** A6 is an umbrella value driver. Its submodules can overlap. Do not simply add every LP submodule at base or upside without checking whether they draw from the same loss pool.

**Use cases:**

- POS/video correlation
- self-checkout anomaly review
- sweethearting detection
- return fraud review
- claims and refund review
- investigation acceleration
- cross-store pattern detection
- dual-use camera analytics
- phantom inventory / on-shelf availability visibility
- inventory and process-control exception review
- safety escalation workflows, treated as Layer C protected value by default

### A6.1 Primary shrink reduction

```text
Shrink Reduction Value =
Annual Revenue
× Shrink Rate
× Addressable Shrink %
× Reduction of Addressable Shrink %
× Deployment %
```

**Default shrink-rate and addressability assumptions:**

| Metric | Conservative | Base | Upside / ceiling | Source support |
|---|---:|---:|---:|---|
| Shrink rate (% of sales) | 1.10% | 1.60% | 2.10% for high-risk segments | `SRC-LP-001`, `SRC-LP-001B` |
| Addressable shrink (% of total shrink) | 30% | 50% | 65%; 73% ceiling with justification | `SRC-LP-002` |
| Reduction of addressable shrink | 5-10% | 10-15% | 15-20% cap in early-stage | `SRC-LP-001`, `SRC-LP-004`; validation from `SRC-LP-005`, `SRC-LP-006`, `SRC-LP-007` |

**Segment note:** Use customer-provided shrink rates first. If the retailer does not provide a rate, use Section 4.8.2. Grocery, pharmacy, mass merchant, electronics, and other high-risk / high-volume formats may justify higher planning assumptions. QSR / restaurant scenarios should not use general retail shrink by default; use a food-waste / process-loss proxy only when explicitly in scope. Always label segment defaults as benchmark-informed.

### A6.2 Return fraud / abuse reduction

```text
Return Fraud Reduction Value =
Annual Return Value
× Fraud / Abuse Rate
× Addressable %
× Reduction %
× Deployment %
```

**Definition:** `Annual Return Value` means the gross dollar value of returned merchandise at retail price. Use customer-provided return value when available. If unknown, estimate as `annual_revenue × return_rate` using the segment-aware return-rate table in Section 4.8.1 and label it as a benchmark fallback.

**v1.11 materiality gate:** If `annual_return_value` is benchmark-derived, suppress the numeric value from the executive Layer A total and mark the driver as **requires customer return data** when either condition is true:

1. return-fraud reduction would exceed **10% of Net Layer A**, or
2. return fraud would become a **top-3 Layer A value driver**.

The agent may present the calculation as an assumptions / sensitivity appendix, but it must not count it in Net Layer A until the customer provides actual return value, return rate, or fraud / abuse baseline. If return fraud becomes a top-5 Layer A driver but does not trigger suppression, still request customer-provided `annual_return_value` before external use.

**Return-fraud planning range:**

| Metric | Conservative | Base | Upside / ceiling | Source support |
|---|---:|---:|---:|---|
| Return fraud / abuse rate (% of returns) | 8.00% | 13.20% | 15.14% | `SRC-LP-003`, with `SRC-LP-002` as broader TRL planning-range support |
| Return fraud / abuse reduction | 5-10% | 10-15% | 15-20% cap in early-stage | `SRC-LP-003`, `SRC-LP-006`, `SRC-LP-007` |

**Overlap rule:** Apply a mandatory 20-30% overlap adjustment between shrink reduction and return fraud reduction unless the retailer can show that return fraud is not already included in the shrink / loss pool being modeled.

### A6.3 Optional self-checkout exposure submodule

Default this submodule **OFF** unless the retailer provides SCO sales mix or explicitly asks for a self-checkout / Scan & Go scenario.

```text
SCO Incremental Loss Exposure =
Annual Revenue
× SCO Sales Mix %
× (SCO Loss Basis Points / 10,000)
× Deployment %
```

**Worked unit example:**

```text
If annual revenue = $5B
SCO sales mix = 30%
SCO loss assumption = 2.6 bps
Deployment = 100%

SCO Incremental Loss Exposure =
$5B × 30% × (2.6 / 10,000) × 100%
= $390,000
```

**SCO planning range:**

| Scenario | Planning assumption | Source support |
|---|---:|---|
| Conservative fixed SCO | 1.0 bps applied to revenue in SCO scope | `SRC-LP-004` |
| Base fixed SCO / benchmark-informed | 2.6 bps applied to revenue in SCO scope | `SRC-LP-004`; validation only from `SRC-LP-005` |
| Upside / higher-risk checkout mix | 5.0 bps cap in early-stage | `SRC-LP-004` |

**Guardrail:** Do not use 7.0+ bps assumptions in early-stage scenarios unless the customer provides Scan & Go / BYOD checkout mix and current loss baseline. `SRC-LP-005` is validation only; do not use it as the default SCO upside anchor.

### A6.4 Investigation productivity

```text
Investigation Savings =
Annual Investigations
× Baseline Hours per Investigation
× Time Saved %
× LP Labor Rate
× Deployment %
```

**Investigation planning range:**

| Metric | Conservative | Base | Upside | Source support |
|---|---:|---:|---:|---|
| Baseline investigation time | Customer-provided; fallback tiered by investigation volume | Customer-provided; fallback tiered by investigation volume | Customer-provided; fallback tiered by investigation volume | `SRC-LP-006`, `SRC-LP-007`; Section 4.8.3 |
| Investigation time saved | 30% | 50% | 70% cap in early-stage | `SRC-LP-006`, `SRC-LP-007` |

**Guardrail:** Investigation savings can coexist with shrink reduction only if it represents LP labor productivity rather than the same prevented loss dollars. If investigation efficiency is being used to justify shrink reduction, apply the A6 investigation × shrink overlap rule. For high-volume LP teams, do not default to 9 hours per case unless the customer confirms that average; use Section 4.8.3 tiering.

### A6.5 Optional inventory / on-shelf availability support

Default this driver **OFF** unless the retailer provides an OSA baseline, affected category sales, BOPIS failure data, stockout data, or a specific inventory-accuracy use case.

```text
OSA / Phantom Inventory Value =
Customer-identified OSA-affected category sales
× OSA Improvement %
× Sales Sensitivity Factor
× Deployment %
```

**Definition:** `Sales in Scope` means only the category, department, store cohort, or omnichannel revenue exposed to measured OSA / phantom inventory issues. Do not use total annual revenue unless the customer explicitly confirms enterprise-wide OSA exposure.

| Metric | Conservative | Base | Upside | Source support |
|---|---:|---:|---:|---|
| OSA sales sensitivity factor | 0.10% | 0.30% | 0.60% cap in early-stage | `SRC-LP-008`, `SRC-LP-009`, `SRC-INV-002`, `SRC-INV-003` |

**Guardrail:** Treat OSA / phantom inventory as optional operational upside, not automatic shrink value. Avoid stacking OSA value on top of shrink or BOPIS recovery unless the populations are clearly separated.

### A6.6 Safety treatment

Safety and workplace-violence sources support **Layer C protected value** and operational-risk framing. They should not be quantified in A6 Layer A by default.

If the user asks to quantify safety, use C4 Workplace Safety Protected Value in Section 8.4 and request customer-specific data such as incident volume, claims cost, workers' compensation, turnover cost, insurance premiums, or legal exposure.

**Cisco relevance:** Meraki cameras, Cisco networking, PoE, Splunk, Webex Connect, Cisco security, AppDynamics, ThousandEyes, Cisco Spaces, and integrated POS/video analytics patterns.

**Source support:** `SRC-LP-001`, `SRC-LP-001B`, `SRC-LP-002`, `SRC-LP-003`, `SRC-LP-004`, `SRC-LP-005`, `SRC-LP-006`, `SRC-LP-007`, `SRC-LP-008`, `SRC-LP-009`, `SRC-LP-010`, `SRC-LP-011`, `SRC-LP-012`, with optional inventory / OSA support from `SRC-INV-001`, `SRC-INV-002`, `SRC-INV-003`, and `SRC-INV-004`.

### A6.7 A6 umbrella cap and guardrails

**A6 umbrella cap:** Total modeled A6 Layer A value should not exceed **60-70% of the customer-confirmed total addressable loss pool** unless the user explicitly provides distinct, non-overlapping loss categories. If the addressable loss pool is benchmark-derived rather than customer-confirmed, use the lower end of the cap and label the output as directional.

**Guardrails:**

- Do not claim Cisco or connected LP eliminates all shrink.
- Distinguish total shrink from addressable / preventable shrink.
- Treat 73% addressable shrink as a ceiling, not a default.
- Do not assume every detected fraud event converts to recovered cash.
- Do not stack return fraud, shrink, SCO loss, and phantom inventory value if they are measuring the same underlying loss pool.
- Treat vendor-commissioned TEI and vendor case studies as range validation / supporting evidence unless the customer has comparable deployment data.
- Keep safety quantification in Layer C unless customer-specific cost data supports a separate finance conversation.
- If more than two A6 submodules are active, explicitly show the overlap controls and umbrella cap used.
- A6 internal submodule overlap adjustments apply before external overlap with other drivers. External A6 overlaps should be applied to **net A6 after internal overlap**, not to each submodule again. A6 internal overlap adjustments must not be used to bypass the A6 umbrella cap.

**Double-counting warning:** Do not count the same loss pool again under inventory accuracy, process control, safety, protected value, or omnichannel experience. If phantom inventory produces both shrink reduction and sales recovery, separate the populations or apply an explicit overlap adjustment.

---

## 6.7 EX Value Driver A7: Intelligent Labor and Service Flow

**Purpose:** Estimate economic value from queue reduction, wait-time improvement, better service flow, pickup / returns coordination, and labor redeployment.

**Use cases:**

- checkout queue detection
- returns counter flow management
- pickup / curbside tasking
- department coverage alerts
- fitting room support
- dwell-based associate intervention
- store manager visibility

**Queue recovery formula:**

```text
Recovered Queue Value =
At-Risk Queue Revenue
× Abandonment Reduction %
× Deployment %
```

**Labor redeployment formula:**

```text
Labor Value =
Hours Redeployed
× Labor Rate
× Realization Factor
```

**Default assumptions:**

| Scenario | Abandonment reduction |
|---|---:|
| Conservative | 5% - 10% |
| Base | 10% - 20% |
| Upside | 20% - 35% |

**Cisco relevance:** Cisco Spaces, Meraki cameras, Wi-Fi, Webex Connect, Splunk.

**Double-counting warning:** Do not stack queue recovery with personalization uplift unless transaction populations are distinct.

---

## 6.8 EX Value Driver A8: AI-Assisted Workflow Protection

**Purpose:** Estimate economic value from reducing routine associate interruptions and low-value repetitive service burden.

**Use cases:**

- associate assist
- routine customer question automation
- order status deflection
- product Q&A support
- return policy guidance
- escalation routing
- post-interaction summaries

**Formula:**

```text
Associate Interruption Savings =
Routine Store Inquiries
× Deflection %
× Minutes Saved per Inquiry
× Labor Rate
× Realization Factor
× Deployment %
```

**Alternative formula:**

```text
Deflected Assisted Contact Value =
Deflected Assisted Contacts
× (Assisted Cost - Self-Service Cost)
```

**Default range:**

| Scenario | Deflection |
|---|---:|
| Conservative | 10% - 15% |
| Base | 15% - 30% |
| Upside | 30% - 50% |

**Cisco relevance:** Webex AI, Webex Connect, Webex Contact Center, Splunk, AppDynamics.

**Double-counting warning:** If the same customer contacts are counted in AI-Powered Service and Outreach, do not count them again here.

**Overlap caution:** A8 is structurally prone to overlap because it can touch customer service, labor flow, and task automation. If A3, A7, and A9 are all active, consider whether A8 should be narrowed to a distinct inquiry population or folded into the other active drivers before calculating Layer A.

---

## 6.9 EX Value Driver A9: Task Automation and ESLs

**Purpose:** Estimate economic value from price-change labor savings, pricing accuracy, promotion execution, and shelf execution.

**Use cases:**

- automated price updates
- promotion execution
- shelf-to-POS price consistency
- pricing error reduction
- shelf-edge product education
- retail media activation
- ESL update monitoring

**Labor formula:**

```text
ESL Labor Savings =
Store Count
× Weekly Price-Change Hours Saved
× 52
× Labor Rate
× Deployment %
× Realization Factor
```

**Pricing error formula:**

```text
Pricing Error Savings =
Store Count
× Annual Pricing Error Cost per Store
× Reduction %
× Deployment %
```

**Default weekly hours saved:**

| Scenario | Hours saved per store per week |
|---|---:|
| Conservative | 4 - 6 |
| Base | 8 - 12 |
| Upside | 12 - 20+ |

**Cisco relevance:** Cisco networking, PoE, Meraki, ESL infrastructure, Splunk, AppDynamics, ThousandEyes.

**Double-counting warning:** Do not stack ESL labor savings with broader task automation unless the labor tasks are distinct.

---

## 6.10 EX Value Driver A10: Connected Facilities and Smart Lighting

**Purpose:** Estimate economic value from lighting energy savings, facilities visibility, occupancy-based control, and connected building operations.

**Use cases:**

- occupancy-based lighting
- daylight harvesting
- task lighting
- safety lighting
- merchandising emphasis
- after-hours monitoring
- facilities telemetry
- PoE-powered smart infrastructure

**Energy formula:**

```text
Smart Lighting Savings =
Annual Utility Spend
× Lighting Share %
× Lighting Energy Savings %
× Deployment %
```

**Default assumptions:**

| Scenario | Lighting energy savings |
|---|---:|
| Conservative | 10% - 15% |
| Base | 15% - 30% |
| Upside | 30% - 50%+ |

**Cisco relevance:** Cisco Catalyst PoE / PoE++, Meraki, sensors, Splunk.

**Double-counting warning:** Apply savings only to lighting-related energy, not total utility spend, unless the user provides broader building-control savings.

---

## 6.11 EX Value Driver A11: Connected Device Reliability

**Purpose:** Estimate economic value from fewer device failures, less associate downtime, fewer tickets, and fewer truck rolls.

**Use cases:**

- handheld performance assurance
- POS and checkout resilience
- camera and sensor uptime
- digital signage assurance
- ESL update monitoring
- incident correlation
- remote troubleshooting
- priority-based remediation

**Productivity formula:**

```text
Device Productivity Savings =
Device-Related Incidents
× Average Lost Time
× Labor Rate
× Deployment %
```

**Support formula:**

```text
Support Savings =
Avoided Tickets × Cost per Ticket
+
Avoided Truck Rolls × Cost per Truck Roll
```

**Default reduction range:**

| Scenario | Incident reduction |
|---|---:|
| Conservative | 10% - 20% |
| Base | 20% - 40% |
| Upside | 40% - 60% |

**Cisco relevance:** Meraki, Catalyst, ThousandEyes, AppDynamics, Splunk, Cisco-Zebra integrations.

**Double-counting warning:** Avoid double-counting with observability MTTR savings if the same incidents are included.

---

## 6.12 IT Value Driver A12: Secure Connected Store Foundation

**Purpose:** Estimate direct IT value from fewer network incidents, lower support burden, and reduced operational fragility.

**Use cases:**

- store Wi-Fi modernization
- cloud-managed network operations
- SD-WAN resilience
- PoE device support
- critical workflow prioritization
- device visibility
- remote troubleshooting
- new use case readiness

**Formula:**

```text
Network Incident Savings =
Annual Network Incidents
× Incident Reduction %
× Cost per Incident
× Deployment %
```

**Support formula:**

```text
Support Savings =
Reduced Tickets × Cost per Ticket
+
Avoided Truck Rolls × Cost per Truck Roll
```

**Default range:**

| Scenario | Incident reduction |
|---|---:|
| Conservative | 10% - 20% |
| Base | 20% - 40% |
| Upside | 40% - 60% |

**Cisco relevance:** Catalyst, Meraki, Wi-Fi, SD-WAN, PoE, Secure Access, ThousandEyes, Splunk.

**Double-counting warning:** Outage exposure belongs mostly in Layer C unless the customer can isolate direct avoided costs.

---

## 6.13 IT Value Driver A13: Observability as a Business Capability

**Purpose:** Estimate direct IT value from MTTR reduction, faster root cause, fewer war rooms, and fewer dispatches.

**Use cases:**

- checkout journey assurance
- pickup and fulfillment monitoring
- digital experience assurance
- store-device incident correlation
- SaaS and internet visibility
- incident prioritization
- proactive remediation

**MTTR formula:**

```text
MTTR Savings =
Annual Incidents
× Current MTTR
× MTTR Reduction %
× IT Labor Cost per Hour
× Deployment %
```

**Truck roll formula:**

```text
Avoided Dispatch Savings =
Avoided Truck Rolls × Cost per Truck Roll
```

**Default MTTR reduction:**

| Scenario | MTTR reduction |
|---|---:|
| Conservative | 10% - 20% |
| Base | 25% - 50% |
| Upside | 50% - 85% |

**Cisco relevance:** Splunk, AppDynamics, ThousandEyes, Meraki, network telemetry.

**Double-counting warning:** Do not double-count the same incidents under secure foundation, observability, and device reliability.

## 6.14 Shared truck-roll allocation rule for A11 / A12 / A13

A single avoided truck-roll or field-dispatch pool may only be allocated once across A11 Connected Device Reliability, A12 Secure Connected Store Foundation, and A13 Observability unless the customer provides distinct truck-roll categories.

**Default allocation rule:**

| Situation | Treatment |
|---|---|
| Customer provides one annual truck-roll number only | Allocate the pool to the strongest matching driver, then mark truck-roll savings as unavailable for the other A11/A12/A13 drivers. |
| Customer provides distinct dispatch categories, such as device, network, application, and carrier / ISP | Allocate each category to the corresponding driver and document the allocation. |
| The agent calculates both MTTR savings and avoided truck-roll savings from the same incident pool | Apply the A11/A12/A13 overlap controls and show the allocation table in the output. |
| Truck-roll savings contributes more than 10% of Net Layer A | Show the truck-roll formula inputs, source of the truck-roll count, cost per truck roll, allocation decision, and overlap adjustment. |

Do not count the same avoided dispatch as device reliability savings, network foundation savings, and observability savings at the same time.

---

# 7. Layer B: First-Party Context Engine Strategic Value Proxy

Layer B is a major component of the Cisco Retail 2.0 TEV. It values the strategic data asset created when connected stores convert anonymous physical behavior into known, consented, CRM-ready, behavior-rich customer context.

> Layer B represents the strategic asset value of the first-party customer context estate. It is not annual cash flow, booked revenue, ROI, or profit.

Layer B is **not booked revenue**. It is a strategic proxy. It may become direct value only through activated and measured programs such as personalization, CPaaS messaging, retail media, loyalty activation, product-intent follow-up, or supplier-funded campaigns.

Layer B includes:

- Wi-Fi / app / loyalty identity capture
- CRM-ready email or phone capture
- Cisco Spaces / location intelligence behavior
- CPaaS / SMS / RCS / WhatsApp messaging identity
- shelf-edge QR / GS1 2D product-intent signals
- zero-party declared preferences
- retail media addressability

## 7.1 Layer B proxy values

Use these ranges unless customer-specific values are available.

| Proxy category | Conservative | Base | Upside |
|---|---:|---:|---:|
| Value per known store session | $0.50 - $1.50 | $2 - $5 | $8 - $15 |
| Value per enriched CRM-ready profile | $3 - $8 | $10 - $30 | $45 - $100 |
| Value per consented messaging identity | $10 - $20 | $30 - $50 | $75 - $120 |
| Value per qualified product-intent signal | $0.30 - $0.80 | $1.50 - $3.00 | $4 - $8 |
| Value per zero-party declared preference | $2 - $5 | $8 - $15 | $20 - $35 |
| Value per retail-media addressable audience member | $1 - $3 | $5 - $15 | $25 - $50 |

## 7.1.1 Layer B output formatting rule

When reporting Layer B, always present it as a **range** and place the disclaimer immediately next to the number.

Required adjacent language:

```text
Layer B is a strategic asset-value proxy for the first-party customer context estate. It is not annual revenue, booked cash flow, ROI, or profit.
```

If total Layer B exceeds 25% of annual revenue, do **not** cap it automatically. Instead, flag it:

```text
Layer B exceeds 25% of annual revenue. This may be plausible for a large first-party data asset, but it must be presented only as strategic asset proxy value and should not be interpreted as annual revenue or savings.
```

If total Layer B exceeds 50% of annual revenue, add an escalation note rather than suppressing the value:

```text
Layer B exceeds 50% of annual revenue. Treat this as an aggressive strategic asset-value scenario. Before using externally, review the identified-session rate, opt-in / match rate, deduplication factor, and proxy-value assumptions. Consider showing conservative and base cases more prominently unless the user explicitly confirms the high identity coverage and activation maturity assumptions.
```

This is a **credibility flag, not a cap**. Layer B remains an asset-value proxy and should not be treated as annual P&L.

## 7.2 B1: Known Store Session Creation

**Formula:**

```text
Known Store Sessions =
Annual Store Visits
× Identified Session Rate
× Deployment %
```

**Value formula:**

```text
Known Store Session Value =
Known Store Sessions
× Value per Known Store Session
```

**Interpretation:** A known store session is a physical visit linked to loyalty, app, Wi-Fi, CRM, email, phone, device, or other consented customer context.

**High-frequency segment guidance:** Apply Section 4.8.5 when the retailer has high visit frequency or many repeat transactions per identified customer. For grocery and pharmacy, use the standard numeric dampening factors. For QSR, fast casual, convenience, fuel / c-store, and similar high-frequency transactional formats, use the stricter v1.11 QSR / convenience override in Section 4.8.5. If active customer count is unknown, use the lower end of the selected B1 range and ask for unique customer / active loyalty data before using externally.

---

## 7.3 B2: Enriched CRM-Ready Profiles

**Formula:**

```text
Enriched CRM Profiles =
Known / Captured Profiles
× Loyalty Active Rate or Reachability Factor
× Enrichment Rate
× Deduplication Factor
```

**Active-user rule:** Use either `total_loyalty_members × loyalty_active_rate` **or** a customer-provided `active_loyalty_or_app_users` count. Do **not** apply `loyalty_active_rate` again when the user already provided active loyalty / app users. If the user provides both total members and active users, prefer active users for B2 and cite total members only as context.

**Value formula:**

```text
CRM Context Value =
Enriched CRM Profiles
× Value per Enriched CRM-Ready Profile
```

**Interpretation:** An enriched CRM-ready profile is more than an email address. It may include loyalty identity, store visit behavior, category interest, dwell signals, shopping missions, product interactions, service needs, preference data, or purchase context. If the user provides total loyalty members rather than active loyalty/app customers, apply `loyalty_active_rate` from Section 4.2 or ask for the active/reachable count.

---

## 7.4 B3: Consented Messaging Identities

**Formula:**

```text
Consented Messaging Identities =
Unique SMS / RCS / WhatsApp / Phone Identities
× Identity Validity Rate
```

**Value formula:**

```text
Messaging Identity Value =
Consented Messaging Identities
× Value per Messaging Identity
```

**Interpretation:** A phone-based or messaging identity is usually more valuable than a generic email record because it supports direct, high-response, conversational engagement. These values assume valid, current, usable consent or channel eligibility. Outbound use must still follow channel, consent, and policy rules.

---

## 7.5 B4: Qualified Product-Intent Signals

**Formula:**

```text
Qualified Product-Intent Signals =
Annual Scans
× Qualified Intent Rate
× Opt-In / Match Rate
```

**Value formula:**

```text
Product Intent Value =
Qualified Product-Intent Signals
× Value per Qualified Product-Intent Signal
```

**Examples of product-intent signals:**

- allergen question
- compatibility question
- recipe request
- coupon interest
- product comparison
- service request
- warranty question
- “where can I find this?”
- “does this work with X?”

---

## 7.6 B5: Zero-Party Declared Preferences

**Formula:**

```text
Zero-Party Declared Preferences =
Preference Captures
× Qualified Preference Rate
```

**Value formula:**

```text
Zero-Party Preference Value =
Zero-Party Declared Preferences
× Value per Declared Preference
```

**Interpretation:** Zero-party data is information a customer intentionally declares, such as preferences, constraints, needs, allergies, style, budget, intended use, or product compatibility requirements.

---

## 7.7 B6: Retail-Media Addressable Audience

**Formula:**

```text
Retail-Media Addressable Audience =
Known Audience Members
× Segmentability Rate
× Media Activation Readiness %
```

**Value formula:**

```text
Audience Value =
Retail-Media Addressable Audience
× Value per Retail-Media Audience Member
```

**Alternative monetized media formula:**

```text
Retail Media Economic Value =
Addressable Impressions / 1,000
× CPM
× Media Value Factor
```

**Guardrail:** Monetized media revenue belongs in Layer A. Addressable audience proxy value belongs in Layer B. Use either the audience proxy model or the monetized media revenue model for the same audience unless the model clearly separates realized media revenue from additional unmonetized audience asset value.

---

## 7.8 Layer B double-counting rules

Layer B can coexist with Layer A because it measures a different value type: strategic asset value rather than current-period economic action value. The control is not “choose A or B.” The control is to avoid counting the same Layer B signal twice.

| Situation | Treatment |
|---|---|
| A scan creates a product-intent signal and later a transaction | Count the transaction or commercial action in Layer A. Count the product-intent signal in Layer B as strategic proxy value. |
| A scan captures a phone/email identity and also captures a product question | Count the identity under B2/B3 and the product question under B4/B5 only if they are distinct captured data assets. |
| A messaging identity drives a conversion | Count conversion in Layer A if measured. Count identity in Layer B separately as strategic proxy. |
| A retail media campaign creates media revenue and product sales | Count supplier-paid media economics separately from gross product sales. Do not count the same campaign audience under B6 if already monetized in A4 unless additional unmonetized audience value is isolated. |
| A CRM profile is already known but becomes enriched | Value only the enrichment, not the full profile capture. |
| A Wi-Fi session is anonymous | Count as store intelligence only, not CRM-ready identity. |
| A Wi-Fi session is linked to loyalty/app/phone/email | Count as known store session and potentially enriched profile. |

---

# 8. Layer C: Protected Value at Risk

Layer C represents value protected by resilience, observability, security, uptime, segmentation, and risk reduction.

Layer C should be shown separately and should not be added to direct annual impact by default.

## 8.1 C1: Downtime and Store Workflow Protection

**Formula:**

```text
Protected Downtime Value =
Downtime Exposure per Hour
× Downtime Hours at Risk
× Risk Reduction %
× Deployment %
```

**v1.8 downtime benchmark treatment:** Customer-provided downtime exposure should be used first. If the user asks for benchmark context, present a tiered selector as Layer C protected-value context only, not as a hard default:

| Downtime exposure tier | Use case | Treatment |
|---|---|---|
| Legacy / conservative lower-bound benchmark | Historical cross-industry downtime references | Supporting context only; do not use as default for modern retail. |
| Enterprise cross-industry benchmark | Broad downtime-risk discussion | Hero / protected-value context; disclose source and scope. |
| Retail / omnichannel peak-hour benchmark | POS, payments, BOPIS, app, loyalty, and cloud workflow interruption | Use only with customer confirmation or explicit scenario framing. |

Do not calculate numeric downtime protected value from benchmark hourly costs unless the user explicitly asks for a benchmark scenario and acknowledges that actual exposure depends on affected stores, peak/off-peak timing, workflow scope, recovery time, and redundancy.

**Alternative formula:**

```text
Protected Revenue / Workflow Value =
Revenue or Productivity Exposure per Hour
× Outage Hours Avoided
× Attribution Factor
```

**Applies to:**

- POS and payments
- pickup / curbside workflows
- loyalty and app redemption
- ESL updates
- digital signage
- cameras and sensors
- associate handhelds
- contact center and customer messaging

**Cisco relevance:** Meraki, Catalyst, SD-WAN, ThousandEyes, AppDynamics, Splunk.

---

## 8.2 C2: Cyber Resilience and Risk Reduction

**Formula:**

```text
Cyber Protected Value =
Cyber Exposure
× Annual Incident Probability
× Risk Reduction %
× Deployment %
```

**Benchmark-exposure guardrail:** Use customer-specific cyber exposure when available. If using benchmark exposure, anchor against published breach-cost ranges such as IBM Cost of a Data Breach and Verizon DBIR (`SRC-CYB-001`, `SRC-CYB-002`). Do not produce exposure values materially above published retail or cross-industry benchmarks without explicit customer-provided justification.

**Applies to:**

- ransomware
- credential theft
- account takeover
- third-party access risk
- edge/VPN exploitation
- shadow AI exposure
- customer data exposure
- payment-adjacent risk
- store IoT / OT risk

**Cisco relevance:** Duo, ISE, Secure Access, Umbrella, XDR, Splunk.

---

## 8.3 C3: Zero Trust and Segmentation Protection

**Formula:**

```text
Segmentation Protected Value =
Estimated Lateral Movement Exposure
× Probability Factor
× Segmentation Risk Reduction %
× Deployment %
```

**Applies to:**

- POS/payment segmentation
- guest Wi-Fi isolation
- IoT segmentation
- camera / signage / ESL segmentation
- vendor access controls
- associate device access
- facilities and smart building systems

**Guardrail:** Do not stack fully with cyber resilience if both use the same breach exposure pool.

## 8.4 C4: Workplace Safety and Associate Risk Protection

Safety value belongs in Layer C by default. It should not be counted as Layer A cash savings unless the customer provides incident costs, turnover costs, workers' compensation costs, claims history, insurance premiums, legal exposure, or other specific financial data.

**Formula only if customer data exists:**

```text
Workplace Safety Protected Value =
Safety Incident Exposure
× Risk Reduction %
× Deployment %
```

**Numeric-output gate:** Do not produce a numeric C4 value unless the user provides at least one customer-specific cost input such as workers' compensation cost, claims history, turnover cost, insurance premium exposure, legal exposure, incident-response cost, or documented safety incident cost. If those inputs are not available, describe C4 qualitatively as protected-value / risk-reduction support.

**Segment-aware C4 prompts:** For pharmacy, convenience, grocery, mass merchant, and other high-foot-traffic or ORC-exposed formats, automatically ask whether the retailer tracks safety-related costs such as associate threats, assaults, workers' compensation, security incidents, claims, turnover linked to safety, insurance premium exposure, or legal costs. Do not quantify unless the customer provides one of these cost inputs.

**Applies to:**

- workplace violence / threatening incidents
- ORC-related associate safety exposure
- de-escalation workflows
- public-view monitor deterrence
- emergency response and escalation
- claims, insurance, workers' compensation, turnover, or legal exposure

**Source support:** `SRC-LP-010`, `SRC-LP-011`, `SRC-LP-012`.

**Guardrail:** Treat safety as protected value and operational risk reduction. Do not imply that every safety incident is prevented or that all protected safety exposure becomes direct savings.

---

## 8.5 Layer C overlap guidance

Layer C values are non-additive by default, but they can still overlap with each other. When presenting Layer C, keep downtime / workflow continuity, cyber resilience, and segmentation protection as separate protected-value views unless the customer can prove that the exposure pools are distinct.

| Layer C overlap pair | Guidance |
|---|---|
| C1 Downtime / workflow continuity + C2 Cyber resilience | Do not count the same outage or incident exposure as both operational downtime and cyber protected value unless the scenarios are distinct. |
| C2 Cyber resilience + C3 Zero Trust / segmentation | Treat segmentation as one mechanism that reduces cyber blast radius. Do not add it on top of the same breach exposure without isolating the exposure pool. |
| C1 Downtime / workflow continuity + C3 Zero Trust / segmentation | If segmentation is being valued as continuity protection for store systems, do not also include the same systems in downtime protection without an adjustment. |
| C4 Workplace safety + A6 Loss prevention | Safety should remain in Layer C unless the customer provides hard cost data. Do not count the same safety workflow as both A6 direct value and C4 protected value. |

**Layer C output rule:** Show Layer C as a protected-value range or separated exposure views. Do not collapse C1, C2, C3, and C4 into a single additive number unless the agent clearly explains the exposure pools and non-overlap assumptions.

---

# 9. Layer D: Future Option Value

Layer D captures strategic optionality from the Cisco AI execution foundation.

Layer D should usually be qualitative unless the customer has a specific roadmap and known avoided costs.

Examples:

- future AI agents for store operations
- expanded computer vision
- future retail media formats
- product-originated AI journeys
- edge AI workloads
- robotics / automation readiness
- smart building integrations
- supplier data products
- agentic commerce
- new store formats

**Optional formula only if data exists:**

```text
Future Option Value =
Avoided Rework Cost
+ Deployment Acceleration Value
+ Avoided Duplicate Infrastructure Cost
```

**Guardrail:** Do not count future AI use cases as current-period value unless activated, contracted, resourced, measured, and governed.

---

# 10. Layer A overlap controls

After calculating gross Layer A, apply overlap adjustments. The goal is not to suppress the TEV story; it is to protect credibility when multiple drivers influence the same customer journey, labor pool, transaction, incident, or asset base.

## 10.1 Overlap-method precedence

There are two approved overlap methods: directional category haircuts and the pairwise matrix. **Do not apply both.**

Use the **category haircut method** when the user provides limited inputs, the agent is producing a fast directional estimate, or fewer than three Layer A drivers are active.

Use the **pairwise matrix method** when three or more Layer A drivers are active, when the user asks for a more defensible adjustment, or when a large estimate needs a clearer explanation of which value pools may overlap.

If the agent cannot apply the pairwise matrix cleanly, use the category haircut method and explain that it is a directional adjustment.

## 10.2 Directional category haircuts

| Area | Suggested haircut |
|---|---:|
| CX overlap across personalization, retail media, scan-to-learn | 10% - 25% |
| EX overlap across labor, queue, workflow protection, ESLs | 5% - 20% |
| IT overlap across observability, device reliability, secure foundation | 10% - 25% |
| Shrink / inventory / LP overlap | 5% - 15% |

## 10.3 Pairwise overlap matrix

Use this matrix when multiple Layer A drivers are active. Apply the higher end of the range when the retailer does not have baseline data to isolate the value pools. Apply the lower end when the retailer can show distinct populations, workflows, or measurement boundaries.

| Driver pair | Suggested overlap adjustment | Why |
|---|---:|---|
| A1 Personalized In-Store Engagement + A2 Omnichannel Orchestration | 10% - 20% | Both may influence the same retained or recovered customer journey. |
| A1 Personalized In-Store Engagement + A4 Dynamic Merchandising / Retail Media | 15% - 30% | Same transaction may be influenced by personalized offer and in-store media. |
| A1 Personalized In-Store Engagement + A5 Shelf-Edge Scan-to-Learn | 10% - 25% | Same customer session may generate both personalization lift and scan-to-action value. |
| A3 AI-Powered Service + A8 AI-Assisted Workflow Protection | 15% - 30% | Both may deflect the same routine inquiry population. |
| A4 Dynamic Merchandising / Retail Media + A5 Shelf-Edge Scan-to-Learn | 10% - 25% | Same in-store media/shelf interaction may influence a commercial action. |
| A6 Connected Loss Prevention + A11 Connected Device Reliability | 5% - 15% | Camera/device uptime may support LP outcomes and operational reliability. |
| A6 Shrink Reduction + A6 Return Fraud Reduction | 20% - 30% | Return fraud / abuse may be included in the broader loss pool unless isolated. |
| A6 Shrink Reduction + A6 SCO Loss Exposure | 25% - 40% | SCO loss is often a subset of total shrink unless retailer tracks it separately. |
| A6 Shrink Reduction + A6 Phantom Inventory / OSA | 20% - 40% | Inventory inaccuracy can be a cause or symptom of measured shrink and lost sales. |
| A6 Investigation Savings + A6 Shrink Reduction | 10% - 20% | Investigation efficiency may be one mechanism driving shrink reduction. |
| A6 Phantom Inventory / OSA + A2 Omnichannel Orchestration | 10% - 20% | OSA / BOPIS recovery and omnichannel journey recovery may target the same failed order population. |
| A7 Intelligent Labor / Service Flow + A8 AI-Assisted Workflow Protection | 10% - 20% | Both may recover associate capacity. |
| A8 AI-Assisted Workflow Protection + A9 Task Automation / ESLs | 15% - 30% | Both may claim recovered associate task time. |
| A12 Secure Connected Store Foundation + A13 Observability | 15% - 30% | Both may reduce the same incident pool or MTTR exposure. |
| A11 Device Reliability + A13 Observability | 15% - 30% | Device incidents may also be included in observability savings. |
| A12 Secure Foundation + A11 Device Reliability | 10% - 20% | Improved network foundation may reduce the same device support events. |

## 10.4 Pairwise aggregation rule

When using the pairwise matrix, do **not** simply sum all applicable pairwise percentages. That can over-correct the model.

Preferred approach:

1. Identify all active driver pairs that appear in the matrix.
2. For each overlapping driver, use the **largest applicable pairwise haircut** for that driver rather than stacking multiple haircuts on the same driver.
3. If a single driver appears in **more than three active overlap pairs**, apply the second-largest applicable haircut as an additional adjustment, capped at a **50% total haircut for that driver**. This is most likely when multiple A6 loss-prevention submodules are active.
4. Apply the haircut to the smaller value in the overlapping pair, or to the portion of the driver that plausibly overlaps if the user has provided boundaries.
5. Sum those driver-level adjustments to produce the total overlap adjustment.
6. Explain which pairs drove the adjustment.

Fast early-stage fallback:

- If the agent cannot perform driver-level pairwise math, use the directional category haircut method from Section 10.2.
- State that the category haircut is a simplified credibility adjustment.
- Do not apply the category haircut and the pairwise matrix together.

## 10.5 Practical early-stage scenario rule

If five or more Layer A drivers are active, apply at least a 15% total Layer A overlap adjustment unless the customer provides measurement boundaries. If more than three drivers are modeled at upside, flag the output as aggressive and recommend a value analyst review before using the estimate externally.

## 10.6 Master Layer A formula

Preferred pairwise method:

```text
Gross Layer A Impact =
Sum of all direct annual value drivers

Overlap Adjustment =
Sum of driver-level pairwise adjustments, using the largest applicable pairwise haircut per overlapping driver and avoiding stacked haircuts on the same driver

Net Layer A Impact =
Gross Layer A Impact - Overlap Adjustment
```

Directional fallback method:

```text
Gross Layer A Impact =
Sum of all direct annual value drivers

Overlap Adjustment =
Gross Layer A Impact × Directional Category Haircut %

Net Layer A Impact =
Gross Layer A Impact - Overlap Adjustment
```

## 10.7 Output requirement

Always show:

- gross Layer A before overlap;
- overlap method used: pairwise matrix or directional category haircut;
- overlap adjustment applied;
- net Layer A after overlap;
- why the adjustment was applied.

**Output consistency rule:** Any executive summary, layer summary, priority subtotal, or top-driver list must use the same final net value after applicable overlap controls. Do not show Gross Layer A in the layer-summary table and label it as after-overlap value. If both gross and net are shown, label them clearly as **Gross Layer A before overlap** and **Net Layer A after overlap**.

---

# 11. Labor realization factors

Labor savings should not automatically equal payroll reduction.

Use a realization factor to distinguish direct avoided cost from reallocated capacity.

```text
Realized Labor Value =
Hours Saved × Labor Rate × Realization Factor
```

Default realization factors:

| Labor category | Default realization factor |
|---|---:|
| Associate task minutes | 10% |
| Associate reallocated capacity | 0% - 20% |
| Overtime / seasonal labor | 50% - 100% |
| Avoided contractor / dispatch labor | 80% - 100% |
| Contact center labor | 20% - 60% |
| LP investigation labor | 25% - 75% |
| IT support labor | 50% - 100% |

Explain that non-realized labor productivity may still create TEV through improved service, capacity, execution, and employee experience, but it should not be described as payroll reduction unless the customer confirms that labor hours convert to hard savings.

---

# 12. Recommended output structure

When producing a TEV estimate, use this structure.

## 12.0 Output consistency rules

Before returning a TEV scenario, the agent must reconcile the output so the executive summary, priority totals, top-driver table, driver-detail tables, overlap table, and layer summaries all use the same numbers.

Required rules:

1. **Layer summary uses Net Layer A.** The headline Layer A value in the executive summary and TEV layer-summary table must be **Net Layer A after overlap**, not Gross Layer A.
2. **Gross and Net stay separate.** If Gross Layer A is shown, it must be labeled as **Gross Layer A before overlap**, followed by overlap adjustment and **Net Layer A after overlap**.
3. **Top drivers match detailed drivers.** The top-value-driver table must use the same final net driver values shown in the detailed driver sections. Do not rank a driver using a pre-overlap or preliminary value if a detailed section later shows a different final value.
4. **Every counted driver must appear in the driver breakdown.** If a driver contributes to a priority subtotal or Layer A total, show the driver value, formula basis, and assumptions in the Layer A by value-driver table. If a driver is not shown in the breakdown, do not include it in the subtotal.
5. **Do not bury unshown defaults inside priority totals.** If a driver is estimated using benchmark defaults, list the default and source status. If the driver lacks the minimum required inputs, keep it qualitative or exclude it.
6. **Layer B headline uses the selected scenario range.** For a base-case run, the headline Layer B number should use the base-case range. Do not combine conservative, base, and upside values into an unlabeled blended headline. Show conservative and upside as separate sensitivity bands when useful.
7. **A6 cap uses modeled addressable loss, not revenue throughput.** Self-checkout sales mix and other revenue throughput figures must be converted into modeled loss exposure before being included in A6 cap logic. Do not treat raw SCO revenue throughput as an addressable loss pool.
8. **Confidence is layered.** Provide one overall confidence score and separate driver-level confidence notes for major high- and low-confidence drivers. Do not let a few weak drivers force the whole scenario to Low if many core inputs were provided; instead, explain which drivers are weak.
9. **Source IDs must match the value category.** Layer B source support should come from first-party data, CDP, messaging, identity, loyalty, CAC/replacement-cost, and retail-media sources. Do not cite inventory/RFID or unrelated sources as Layer B proof unless the value being discussed is actually inventory or shelf intelligence.
10. **Final self-check must compare all totals.** Before finalizing, compare the executive Layer A number, Layer A by priority, Layer A by driver, top-driver list, and overlap-adjusted net value. If any mismatch exists, correct it before presenting the output.

## 12.1 Executive TEV summary

```text
Based on the inputs provided, the Cisco Retail 2.0 Connected Store Operating Model produces the following benchmark-informed TEV envelope:

Layer A: $X Net Layer A after overlap, modeled direct annual economic impact
Layer B: $Y-$Y First-Party Context Engine strategic asset-value proxy range for the selected scenario
Layer C: $Z protected value at risk, shown separately and non-additive
Layer D: Future option value potential, described qualitatively unless activated and measured

Figures represent steady-state annual TEV at the deployment scope specified, not Year 1 ramp or implementation timing.
Layer A is the modeled current-period annual impact after overlap controls. Layer B is a strategic asset-value proxy, not annual revenue, booked cash flow, ROI, or profit. Layers B, C, and D are shown separately and should not be treated as booked revenue, guaranteed savings, or ROI.
```

## 12.2 Layer A by priority

| Priority | Direct annual economic impact |
|---|---:|
| CX Growth and Loyalty | $X |
| EX / Operations / Margin Protection | $Y |
| IT Direct Mitigation | $Z |
| Gross Layer A | $A |
| Overlap adjustment | -$B |
| Net Layer A | $C |

## 12.3 Layer A by value driver

**Driver-detail rule:** Every driver included in a priority subtotal must appear in the driver breakdown. Any driver contributing more than **10% of Net Layer A** must show the formula inputs used, which inputs were customer-provided, which were benchmark defaults, and which overlap adjustment was applied. If a driver over 10% of Net Layer A is based primarily on benchmark-derived inputs, label it as a priority discovery item. If that driver is return fraud and `annual_return_value` is benchmark-derived, apply the Section 4.8.1 / A6.2 materiality gate before counting it in Net Layer A.

Every value driver included in a priority subtotal must appear in this table. Values should reconcile to the priority totals and to Net Layer A after overlap. If a value driver is excluded because inputs are missing or gating rules keep it off, mark it as **Excluded / qualitative** and do not include it in the subtotal.

| Value driver | Modeled value | Status / assumption basis |
|---|---:|---|
| Personalized In-Store Engagement | $X | Provided / default / excluded |
| Omnichannel Journey Orchestration | $X | Provided / default / excluded |
| AI-Powered Service and Outreach | $X | Provided / default / excluded |
| Dynamic Digital Merchandising and Retail Media | $X | Provided / default / excluded |
| Shelf-Edge Scan-to-Learn | $X | Provided / default / excluded |
| Margin Protection and Connected Loss Prevention | $X | Provided / default / excluded |
| Intelligent Labor and Service Flow | $X | Provided / default / excluded |
| AI-Assisted Workflow Protection | $X | Provided / default / excluded |
| Task Automation and ESLs | $X | Provided / default / excluded |
| Connected Facilities and Smart Lighting | $X | Provided / default / excluded |
| Connected Device Reliability | $X | Provided / default / excluded |
| Secure Connected Store Foundation | $X | Provided / default / excluded |
| Observability as a Business Capability | $X | Provided / default / excluded |

## 12.4 Layer B summary

Always show Layer B as a range and place the asset-value disclaimer adjacent to the number.

| Component | Strategic proxy value range |
|---|---:|
| Known store sessions | $X |
| Enriched CRM-ready profiles | $Y |
| Consented messaging identities | $Z |
| Product-intent signals | $A |
| Zero-party declared preferences | $B |
| Retail-media addressable audience | $C |
| Total Layer B proxy | $D - $D |

## 12.5 Layer C summary

| Protected value area | Protected value |
|---|---:|
| Downtime / workflow continuity | $X |
| Cyber resilience | $Y |
| Zero trust / segmentation | $Z |
| Workplace safety / associate risk protection | $W |
| Total Layer C protected value | $A |

## 12.6 Top value drivers

Top-driver values must match the final net driver values shown in Section 12.3. Do not rank a driver using gross value if the detailed section shows net value after overlap.

List the top 5 modeled drivers by value and explain why they matter. Use the same final values shown in the detailed driver breakdown. If overlap was applied at the driver level, use the post-overlap value for ranking or clearly label whether the ranking is pre-overlap.

Example:

1. Margin Protection and Connected Loss Prevention
2. Dynamic Digital Merchandising and Retail Media
3. AI-Powered Service and Outreach
4. Task Automation and ESLs
5. Observability as a Business Capability

## 12.7 Assumptions and caveats

Always include:

- customer-provided inputs;
- benchmark defaults used;
- deployment scope;
- scenario selected;
- overlap haircut;
- confidence level;
- what was excluded;
- what not to add together;
- next discovery questions.

---

# 13. Confidence scoring

Each TEV output should include an overall confidence rating and driver-level confidence notes.

**Overall confidence** should reflect the completeness of the customer inputs and the share of the estimate driven by benchmark defaults.

**Driver-level confidence** should identify which value drivers are high confidence, medium confidence, or low confidence. A scenario may have Medium overall confidence even if specific drivers, such as Layer C or retail media, are low confidence.

| Confidence | Criteria |
|---|---|
| High | Customer provided operational inputs, use-case scope, baseline metrics, and deployment assumptions. |
| Medium | Customer provided revenue, stores, and some operating data, but defaults were used for several assumptions. |
| Low | Estimate is mostly benchmark-driven and should be treated as directional planning only. |

Also list the highest-leverage missing inputs.

---

# 14. Recommended discovery questions

After producing the first TEV estimate, ask only the most relevant follow-ups.

## CX discovery

- What percentage of store visits are tied to loyalty, app, Wi-Fi, phone, email, or CRM identity today?
- How many loyalty members and active app users do you have?
- What is your annual BOPIS / pickup volume?
- What is your current customer service contact volume?
- Do you currently use SMS, RCS, WhatsApp, or app messaging for service or marketing?
- Do you have digital signage or in-store media inventory?
- Are shelf-edge QR / GS1 2D / product scan experiences in scope?

## EX / operations discovery

- What is your shrink rate?
- What portion of shrink is concentrated in categories, self-checkout, returns, or specific stores?
- How many hours per week do stores spend on price changes / paper labels?
- Do you use ESLs today?
- What are the biggest store labor friction points?
- Do you have queue or wait-time data?
- What is annual energy spend and lighting share?
- How often do handhelds, scanners, POS, cameras, signage, or ESLs fail?

## IT / security discovery

- How many store-affecting IT incidents occur annually?
- What is current MTTR?
- How many truck rolls or field dispatches happen per year?
- What store systems are most downtime-sensitive?
- What is current network segmentation maturity?
- Are IoT, cameras, signage, ESLs, guest Wi-Fi, POS, and associate devices segmented?
- Do you have a quantified cyber exposure or risk model?
- Do you currently use observability across app, network, cloud, SaaS, and device layers?

---

# 15. Benchmark Source Registry and Source-to-Impact Map

## 15.0 Source governance rules for agents

This file is intended to be portable into any conversational agent. When a user asks where a benchmark came from, the agent must cite the **original external source** listed in this section, not any internal research compilation document used to assemble this calculator.

**v1.4 source-governance update:** This section resolves previously dangling `SRC-CSCO-*` IDs, replaces generic Cisco home-page references with specific product / proof-point pages, demotes or removes weak sources flagged by red-team review, adds source-status labels, and adds a Layer B triangulation methodology.

**v1.6 source-use update:** Manual-verification-required sources should not anchor default formula ranges. Vendor-commissioned TEI sources may validate plausibility but should not be treated as universal defaults without customer-specific evidence.

**Important verification note:** Link and claim status labels below reflect the current source-audit process and should still be manually checked before external distribution, proposal use, legal / brand review, or formal value-engineering work. Some reputable sources may return 403 / 406 / 503 in automated checks because of bot protection; do not treat bot-blocking alone as a failed source.

### Source status labels

| Status | Meaning | Use rule |
|---|---|---|
| **Verified primary** | Original / official source or recognized institutional source. Link and claim have been audit-reported as usable or are source-specific enough for manual verification. | Usable in source registry and customer-facing source support. |
| **Verified vendor-commissioned** | Analyst-modeled TEI or similar source commissioned by a vendor. | Usable with disclosure. Do not present as universal independent evidence. |
| **Verified supporting** | Useful company, trade, standards, or vendor source that supports the concept or planning range. | Use as supporting evidence or planning range, not sole proof for aggressive claims. |
| **Manual verification required** | Link may be bot-blocked, paywalled, unstable, or not fully claim-verified. | Keep in registry only if useful, but flag before external use. |
| **Replace / do not use as primary** | Source is weak, indirect, broken, generic, or better replaced by a primary source. | Do not use as formula support. Replace or demote to appendix. |
| **Remove** | Benchmark not found, contradicted, or unsuitable. | Do not use in model outputs. |

### Source confidence labels

| Confidence | Meaning |
|---|---|
| **High** | Analyst study, primary company source, standards body, recognized institutional source, public retailer / platform proof point, or major security / industry report. |
| **Medium** | Credible vendor, platform, industry publication, or case-study source useful for planning ranges but requiring careful wording. |
| **Directional** | Useful context only; not strong enough alone for formula assumptions. |

### Model role labels

| Role | Meaning |
|---|---|
| **Formula input** | Can support a default benchmark or planning range when customer data is unavailable. |
| **Planning range** | Supports conservative / base / upside ranges, but customer data should replace it when available. |
| **Hero benchmark** | Useful for executive context, but should not be applied mechanically as a formula assumption. |
| **Supporting evidence** | Supports the business rationale for a value driver. |
| **Cisco relevance** | Supports Cisco portfolio fit, but is not necessarily an external benchmark. |
| **Caveat / compliance context** | Should inform guardrails and assumptions, not value creation. |

---

### v1.4.1 lookup-resolution note

This version resolves the remaining v1.4 lookup failures from the source-to-impact map:

- `SRC-INFRA-001` references have been replaced with `SRC-CSCO-002` where the intended support is Cisco Store / Cisco-on-Cisco facilities, lighting, and analytics evidence.
- `SRC-TEV-004` references have been replaced with `SRC-OPS-008` where the intended support is the Zendesk / Forrester TEI customer-service automation source.
- The former `SRC-RM-005` slot is intentionally retired / skipped. Use `SRC-RM-001`, `SRC-RM-001B`, `SRC-RM-002`, `SRC-RM-003`, `SRC-RM-004`, and `SRC-RM-006` for retail-media support. Do not infer a missing active source from the numbering gap.

## 15.1 Source-to-impact map

| Value driver | Impact metric | Preferred source IDs | Use in model |
|---|---|---|---|
| **A1 Personalized In-Store Engagement** | Identified sessions, loyalty engagement, contextual offers | `SRC-FPD-001`, `SRC-FPD-002`, `SRC-FPD-004`, `SRC-FPD-005`, `SRC-FPD-006`, `SRC-AI-001`, `SRC-AI-002`, `SRC-CSCO-001` | Supports A1 identity-driven engagement and B1/B2 known-session/profile enrichment assumptions. |
| **A2 Omnichannel Journey Orchestration** | BOPIS, pickup completion, service recovery, journey recovery | `SRC-AI-001`, `SRC-AI-002`, `SRC-OPS-001`, `SRC-OPS-002`, `SRC-FPD-002` | Supports A2 recovered-journey framing and omnichannel execution rationale. |
| **A3 AI-Powered Service and Outreach** | Contact deflection, SMS/RCS/WhatsApp engagement, conversational conversion, cost-per-resolution | `SRC-FPD-003`, `SRC-CSCO-003`, `SRC-OPS-008`, `SRC-OPS-009` | Supports A3 service-outreach, messaging, contact-deflection, and cost-per-resolution assumptions. `SRC-OPS-008` is vendor-commissioned supporting evidence; `SRC-OPS-009` is a secondary source reporting a Gartner-style self-service versus assisted-service benchmark and should be replaced by customer-specific cost data when available. Do not use removed back-in-stock RPS claims unless reverified. |
| **A4 Dynamic Digital Merchandising and Retail Media** | CPM, supplier-funded media, in-store ad performance, addressable impressions | `SRC-RM-001`, `SRC-RM-002`, `SRC-RM-003`, `SRC-RM-004`, `SRC-RM-006` | Supports A4 monetized media revenue and B6 audience-addressability proxy. `SRC-RM-007` is retired as primary support. |
| **A5 Shelf-Edge Scan-to-Learn** | Product-intent signals, zero-party preferences, scan-to-action value | `SRC-FPD-009`, `SRC-FPD-010`, `SRC-FPD-014`, `SRC-SHELF-001`, `SRC-SHELF-002`, `SRC-RM-001` | Supports A5 scan-to-action value and B4/B5 signal-value assumptions. |
| **A6 Margin Protection and Connected Loss Prevention** | Shrink reduction, preventable shrink, return fraud / abuse, SCO exposure, investigation efficiency, POS-video correlation, phantom inventory, safety | `SRC-LP-001`, `SRC-LP-001B`, `SRC-LP-002`, `SRC-LP-003`, `SRC-LP-004`, `SRC-LP-005`, `SRC-LP-006`, `SRC-LP-007`, `SRC-LP-008`, `SRC-LP-009`, `SRC-LP-010`, `SRC-LP-011`, `SRC-LP-012`, `SRC-INV-002`, `SRC-INV-003` | Supports A6 shrink baseline, preventable / addressable loss pools, return fraud, SCO exposure, AI video / POS correlation, investigation labor, inventory accuracy / OSA, and safety-related protected-value assumptions. `SRC-LP-002` is now verified by the v1.8 source audit, but the 73% preventable-shrink figure remains a ceiling, not a default. |
| **A7 Intelligent Labor and Service Flow** | Queue reduction, wait-time reduction, abandonment reduction, labor redeployment | `SRC-OPS-001`, `SRC-OPS-002`, `SRC-CSCO-001`, `SRC-CSCO-002` | Supports A7 service-flow and queue-friction assumptions. Obscure / wrong-domain queue sources from earlier drafts are retired. |
| **A8 AI-Assisted Workflow Protection** | Associate interruption reduction, routine inquiry deflection, frontline AI | `SRC-OPS-001`, `SRC-OPS-002`, `SRC-FPD-003`, `SRC-CSCO-003`, `SRC-OPS-008`, `SRC-OPS-009` | Supports A8 associate workflow, AI-assist, and self-service assumptions. Use `SRC-OPS-008` / `SRC-OPS-009` with source-type caveats and prefer customer-provided contact-resolution cost where available. |
| **A9 Task Automation and ESLs** | Price-change labor, pricing error reduction, promotion execution | `SRC-OPS-003`, `SRC-OPS-004`, `SRC-OPS-005`, `SRC-SHELF-001`, `SRC-RM-001` | Supports A9 ESL labor and pricing-execution assumptions. Vendor sources should be treated as planning-range support. |
| **A10 Connected Facilities and Smart Lighting** | Lighting energy savings, facilities telemetry, PoE lighting | `SRC-CSCO-002`, `SRC-CSCO-004`, `SRC-TEV-006` | Supports A10 lighting and facilities assumptions. Cisco-owned proof points should be labeled as Cisco-owned. |
| **A11 Connected Device Reliability** | Device downtime, support tickets, truck rolls, endpoint visibility | `SRC-OPS-001`, `SRC-IT-002`, `SRC-IT-003`, `SRC-IT-004`, `SRC-TEV-006` | Supports A11 device reliability, remote troubleshooting, and endpoint visibility. |
| **A12 Secure Connected Store Foundation** | Network incidents, store uptime, workflow continuity, SD-WAN / Meraki value | `SRC-TEV-006`, `SRC-IT-002`, `SRC-IT-003`, `SRC-IT-004`, `SRC-CSCO-004`, `SRC-CYB-006` | Supports A12 foundation and direct IT mitigation assumptions. |
| **A13 Observability as a Business Capability** | MTTR, incident labor, root-cause reduction, downtime visibility | `SRC-TEV-005`, `SRC-IT-001`, `SRC-IT-002`, `SRC-IT-003`, `SRC-IT-004` | Supports A13 observability, MTTR, and incident-response assumptions. `SRC-IT-001` is legacy / hero context only unless the customer confirms a downtime-exposure scenario. |
| **B1-B6 First-Party Context Engine** | Known sessions, CRM-ready profiles, messaging identities, intent signals, audience addressability | `SRC-FPD-001`-`SRC-FPD-014`, `SRC-RM-001`, `SRC-RM-001B`, `SRC-RM-002`, `SRC-RM-003`, `SRC-RM-004`, and `SRC-RM-006`, `SRC-SHELF-001`, `SRC-SHELF-002` | Supports Layer B strategic asset proxy. Values are triangulated, not copied from a single source. See Section 15.7. |
| **C1 Downtime and Workflow Protection** | Downtime exposure, protected workflow value | `SRC-IT-001`, `SRC-IT-002`, `SRC-IT-003`, `SRC-TEV-005`, `SRC-TEV-006` | Supports C1 protected value at risk. Downtime benchmarks are tiered context only; customer-provided exposure should drive numeric calculations. |
| **C2 Cyber Resilience** | Breach exposure, ransomware, credential attacks, shadow AI | `SRC-CYB-001`, `SRC-CYB-002`, `SRC-CYB-003`, `SRC-CYB-004`, `SRC-CYB-007` | Supports C2 cyber protected value. Customer-specific cyber exposure should replace benchmarks. |
| **C3 Zero Trust and Segmentation** | Segmentation, lateral movement, identity risk, IoT isolation | `SRC-CYB-002`, `SRC-CYB-004`, `SRC-CYB-005`, `SRC-CYB-006`, `SRC-CYB-007`, `SRC-CSCO-005` | Supports C3 segmentation protected value. Generic Cisco homepage citations are retired. |
| **D Future Option Value** | Agentic commerce, AI execution foundation, future store capabilities | `SRC-AI-001`, `SRC-AI-002`, `SRC-AI-003`, `SRC-AI-004`, `SRC-TEV-007` | Supports D qualitative future option value. Use “agentic commerce” carefully as a future-option theme, not current-period impact. |

---

## 15.2 Detailed Benchmark Source Registry

### 15.2.1 First-party data, CRM, loyalty, messaging, and Layer B sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-FPD-001` | Forrester Consulting / Amperity | *The Total Economic Impact of Amperity Customer Data Platform* | 2026 | https://amperity.com/resources/amperity-forrester-tei-study | 505% ROI; $13.02M PV benefits; $10.87M NPV; benefits across effective messaging, paid media targeting, data deduplication, and productivity. | B2, B6, A1 | Hero benchmark / supporting evidence | High | Verified vendor-commissioned | Replaces G2-hosted PDF. Disclose as vendor-commissioned TEI. |
| `SRC-FPD-002` | Forrester Consulting / Twilio Segment | *The Total Economic Impact of Twilio Segment* | 2026 | https://tei.forrester.com/go/twilio/segment/?lang=en-us | 15% mobile conversion-rate lift; risk-adjusted profit / productivity benefits from unified customer data activation. | A1, A2, B2 | Formula input / supporting evidence | High | Verified vendor-commissioned | TEI gateway may be campaign-hosted; disclose as vendor-commissioned. |
| `SRC-FPD-003` | Forrester Consulting / Charles | *The Total Economic Impact of Charles Conversational Commerce Platform* | 2026 | https://tei.forrester.com/go/charles/charlestei/ | 249% ROI; EUR1.2M benefits; EUR820K NPV; 550% engagement lift; conversion-rate lift from 3.3% to 12.2%. | A3, B3, A8 | Hero benchmark / planning range | High | Verified vendor-commissioned | Conversational commerce / WhatsApp-specific. Do not generalize to all messaging. |
| `SRC-FPD-004` | Harvard Business Review / Reichheld & Sasser | *Zero Defections: Quality Comes to Services* | 1990 | https://hbr.org/1990/09/zero-defections-quality-comes-to-services | Original source family for the retention/profit relationship often summarized as 5% retention improvement driving 25%-95% profit lift. | A1, B2 | Supporting evidence | High | Verified primary | Use as historical strategic support, not formula input. |
| `SRC-FPD-005` | Bain & Company | *Zero Defections: Quality Comes to Services* | 1990 / hosted by Bain | https://www.bain.com/insights/zero-defections-quality-comes-to-services-harvard-business-review-hbr/ | Bain-hosted page for Reichheld / Sasser retention economics. | A1, B2 | Supporting evidence | High | Verified primary / backup | Use instead of third-hand vendor blog citation. |
| `SRC-FPD-006` | Rivo | *28 Tiered Loyalty Program Statistics Every Ecommerce Brand Should Know* | 2025 | https://www.rivo.io/blog/tiered-loyalty-program-statistics | Loyalty member spend / redeemer / VIP behavior benchmarks. | A1, B1, B2 | Planning range / supporting evidence | Medium | Verified supporting | Vendor/blog source. Use directionally and pair with primary or customer data. |
| `SRC-FPD-007` | Digital Applied | *SMS Marketing Statistics 2026: Open and CTR Data* | 2026 | https://www.digitalapplied.com/blog/sms-marketing-statistics-2026-open-ctr-data | SMS / RCS performance benchmarks. | A3, B3, B4 | Planning range | Medium | Manual verification required | The specific $3.10 back-in-stock RPS benchmark is retired until independently verified. Do not use that exact claim in outputs. |
| `SRC-FPD-008` | MessageFlow | *SMS Marketing Benchmarks 2026: CTR, Open Rates by Industry* | 2026 | https://messageflow.com/blog/sms-marketing-benchmarks/ | SMS engagement and CTR benchmarks by industry. | A3, B3 | Planning range | Medium | Manual verification required | Use as messaging benchmark context only. |
| `SRC-FPD-009` | BlueConic | *Quiz Marketing: The Interactive Marketing Strategy That Converts* | 2025 | https://www.blueconic.com/resources/quiz-marketing | Marmot / quiz proof point; conversion and AOV lift context for zero-party data. | A5, B4, B5 | Hero benchmark / supporting evidence | Medium | Verified supporting | Vendor case study. Do not generalize a single brand result as universal. |
| `SRC-FPD-010` | Octane AI | *The Complete Guide to Product Quizzes* | 2025 | https://www.octaneai.com/resources/quiz-playbook | Product quiz / zero-party data opt-in, conversion, and CLV benchmarks. | A5, B5 | Planning range | Medium | Manual verification required | Vendor source; use for directional zero-party data economics. |
| `SRC-FPD-011` | Prospeo | *Cost Per Acquisition by Industry: 2026 Benchmarks* | 2026 | https://prospeo.io/s/cost-per-acquisition-by-industry | Google / Meta / platform CPA benchmarks used for replacement-cost proxy. | B2, B3 | Planning range | Medium | Manual verification required | Replacement-cost support only. Replace with customer CAC where available. |
| `SRC-FPD-012` | Digital Applied | *Mobile Marketing Statistics 2026: Data Points* | 2026 | https://www.digitalapplied.com/blog/mobile-marketing-statistics-2026-data-points | Shopping app CPI and Day-30 retention inputs for retained-user acquisition cost proxy. | B2, B3 | Planning range | Medium | Manual verification required | Use only for replacement-cost / retained-user proxy, not direct revenue. |
| `SRC-FPD-013` | McKinsey | *The Next Frontier of Personalized Marketing* | 2021 / active source | https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/unlocking-the-next-frontier-of-personalized-marketing | Personalization impact on revenue and marketing-spend efficiency. | A1, B2, B5 | Supporting evidence / hero benchmark | High | Manual verification required | Strong source family; do not apply broad lifts mechanically. |
| `SRC-FPD-014` | Braze | *Zero-Party Data: The Key to Privacy-First Personalization* | 2024 | https://www.braze.com/resources/articles/what-is-zero-party-data | Zero-party data definition, consented preference capture, personalization context. | B5, A5 | Supporting evidence | Medium | Verified supporting | Useful for definitions and business rationale, not dollar-value formula. |
| `SRC-FPD-015` | Academic / PMC | *Zero-party data between hype and hope* | 2022 | https://pmc.ncbi.nlm.nih.gov/articles/PMC9469730/ | Academic caveats on zero-party data, response bias, and limitations. | B5 caveats | Caveat / compliance context | Medium | Demote to appendix | Use as caution, not value support. |

### 15.2.2 Retail media, digital signage, shelf-edge, GS1 2D, and ESL sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-RM-001` | IAB | *In-Store Retail Media: Definitions and Measurement Standards* | 2025 | https://www.iab.com/guidelines/in-store-retail-media/ | In-store retail media definitions and measurement standards. | A4, B6, A5 | Standards / supporting evidence | High | Verified primary | Use for methodology and measurement definitions, not CPM values. |
| `SRC-RM-001B` | IAB Europe | *IAB Europe and IAB In-Store Retail Media Definitions and Measurement Standards* | 2025 | https://iabeurope.eu/knowledge_hub/iab-europe-and-iab-in-store-retail-media-definitions-and-measurement-standards/ | European IAB mirror / supporting standards context. | A4, B6 | Standards / supporting evidence | High | Verified primary | Use as standards support. |
| `SRC-RM-002` | Walmart Connect | *Store Ads with Walmart Connect* | 2025 | https://www.walmartconnect.com/solutions/store-ads | Walmart in-store ad reach and store media proof points. | A4, B6 | Hero benchmark / supporting evidence | High | Verified primary | Primary retailer/platform source. Use for Walmart-specific reach and proof points. |
| `SRC-RM-003` | Modern Retail / Albertsons / Stratacache Walkbase | *Albertsons is putting digital screens for ads in more than a third of its stores* | 2026 | https://www.modernretail.co/technology/albertsons-is-putting-digital-screens-for-ads-in-more-than-a-third-of-its-stores/ | Albertsons in-store screens, Walkbase passive sensor measurement, 60+ control factors, sales-lift proof point. | A4, B6 | Hero benchmark / supporting evidence | High | Verified supporting | Trade publication; strong proof point but not Cisco-specific. |
| `SRC-RM-004` | Grocery TV / PRNewswire | *Grocery TV’s In-Store Retail Media Platform Surpasses 6,500 Stores* | 2025 | https://www.prnewswire.com/news-releases/grocery-tvs-in-store-retail-media-platform-surpasses-6-500-stores-as-company-unveils-new-brand-identity-302571486.html | Grocery TV scale and in-store media platform reach. | A4, B6 | Supporting evidence | Medium | Verified supporting | Press release / company source; useful for scale, not neutral benchmark. |
| `SRC-RM-006` | Walmart Connect | *Cracking the Code to Incremental Growth* | 2025 | https://www.walmartconnect.com/resources/case-studies/2025/cracking-the-code-to-incremental-growth | Walmart supplier / in-store campaign context and incrementality. | A4 | Supporting evidence / hero benchmark | High | Verified primary | Official Walmart Connect case-study page. |
| `SRC-RM-007` | Osmos.ai | *In-Store Retail Media / RMN Monetization Sources* | 2026 | https://www.osmos.ai/blog/retail-instore-media-digital-screens-2026 | Previously used for CPM / margin ranges such as $10-$50 CPM and 50%-70% in-store margin. | A4, B6 | Retired / replace | Medium | Replace / do not use as primary | Audit reported connection refused / unstable link. Do not use as primary source until manually recovered and verified. |
| `SRC-SHELF-001` | GS1 US | *GS1 Digital Link* | 2025 | https://www.gs1us.org/industries-and-insights/by-topic/gs1-digital-link | GS1 Digital Link context for SKU-level digital experiences. | A5, B4, B5 | Supporting evidence | High | Manual verification required | Use for standard / capability context, not value assumption. |
| `SRC-SHELF-002` | GS1 Global | *GS1 Digital Link* | 2025 | https://www.gs1.org/standards/gs1-digital-link | Global GS1 Digital Link standard context. | A5, B4, B5 | Supporting evidence | High | Manual verification required | Use for standard context. |

### 15.2.2B Inventory accuracy, RFID, and on-shelf availability sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-INV-001` | GS1 US / Auburn University RFID Lab | *EPC/RFID Retail Supply Chain Data Exchange Study - Executive Summary* | 2026 / active source | https://documents.gs1us.org/adobe/assets/deliver/urn%3Aaaid%3Aaem%3Aa8d86484-ffce-498d-8c59-fc1223437d58/Case-Study-Executive-Summary-EPC-RFID-Retail-Supply-Chain-Data-Exchange-Study.pdf | RFID / EPC supply-chain data exchange, order accuracy, and supply-chain visibility context. | A6, A11, D | Hero benchmark / supporting evidence | High | Verified primary / study summary | Use only when the customer is explicitly evaluating RFID / inventory-accuracy transformation. Cisco supports connectivity and data movement; do not imply Cisco alone creates RFID accuracy gains. |
| `SRC-INV-002` | ECR Retail Loss Group | *Defining and Assessing Inventory Record Inaccuracy Metrics* | 2024 / active source | https://ecrloss.com/research-paper/inventory-record-inaccuracy-metrics/ | Inventory record inaccuracy definitions and measurement methodology. | A6, A2 | Supporting evidence | High | Verified primary | Use to define inventory-record inaccuracy / phantom inventory; not a default value assumption. |
| `SRC-INV-003` | ECR Retail Loss Group | *Measuring the Sales Impact of Improving Inventory Records* | 2024 / active source | https://ecrloss.com/research-paper/improving-inventory-records/ | Sales-impact relationship from correcting inventory records; supports optional OSA / inventory accuracy scenarios. | A6, A2 | Supporting evidence / optional scoped scenario | High | Verified primary | Use only with scoped `Sales in Scope` and customer OSA / inventory-accuracy baseline. Do not apply broad sales-lift values to total revenue by default. |
| `SRC-INV-004` | Auburn University RFID Lab | *Inventory Accuracy Whitepaper* | 2024 / active source | https://rfid.auburn.edu/papers/documents/InventoryAccuracyWhitepaper.pdf | Inventory accuracy methodology and RFID-related inventory measurement context. | A6, A11, D | Supporting evidence | High | Verified primary | Use for inventory-accuracy methodology; not direct Cisco-attributable economic value. |

### 15.2.3 Store operations, labor, queue, frontline AI, ESL, and productivity sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-OPS-001` | Zebra Technologies | *17th Annual Global Shopper Study* | 2024 | https://www.zebra.com/content/dam/zebra_dam/en/reports/vision-study/retail-report-vision-study-unlocking-excellence-en-us.pdf | Associate technology, shopper / associate expectations, store-device and frontline enablement context. | A2, A7, A8, A11 | Supporting evidence / planning range | High | Verified supporting | Vendor research, but widely used retail operations source. |
| `SRC-OPS-002` | Zebra Technologies / market coverage | *Zebra Global Shopper Study coverage* | 2025 | https://finance.yahoo.com/news/zebra-study-87-retailers-believe-130000219.html | AI and retail associate / operations study coverage. | A2, A7, A8 | Supporting evidence | Medium | Verified supporting | Use when exact 2025 stat is needed; prefer Zebra primary PDF when available. |
| `SRC-OPS-003` | U.S. Chamber of Commerce | *How Digital Shelf Labels Improve Retail Sales* | 2024 | https://www.uschamber.com/co/good-company/launch-pad/digital-shelf-labels-in-retail | ESL examples, manual tag labor, Walmart rollout context. | A9 | Supporting evidence | Medium | Verified supporting | Third-party article; use as example, not universal benchmark. |
| `SRC-OPS-004` | ComQi | *Cost Savings and ROI: The Financial Impact of Electronic Shelf Labels* | 2025 | https://comqi.com/esl-roi/ | ESL labor / pricing execution value. | A9 | Planning range | Medium | Verified supporting | Vendor source; mark as planning range only. |
| `SRC-OPS-005` | Neuroshop | *Electronic Shelf Labels Cost and ROI* | 2025 | https://neuroshop.tech/blog/electronic-shelf-labels-cost-and-roi | ESL relabeling labor / SKU-heavy store assumptions. | A9 | Planning range | Medium | Verified supporting | Vendor source; use cautiously. |
| `SRC-OPS-006` | MRI Software queue source from earlier registry | Queue / wait-time benchmark | 2025 | N/A | Earlier draft queue source. | A7 | Retired | Directional | Replace / do not use as primary | Red-team flagged as wrong-domain / weak source. Replace with stronger queue, NRF, Zebra, or customer data. |
| `SRC-OPS-007` | ScanQueue source from earlier registry | Queue / wait-time benchmark | 2025 | N/A | Earlier draft queue source. | A7 | Retired | Directional | Replace / do not use as primary | Red-team flagged as obscure / weak source. Replace before external use. |
| `SRC-OPS-008` | Zendesk / Forrester Consulting | *The Total Economic Impact of Zendesk* | 2025 | https://tei.forrester.com/go/zendesk/advancedaisupport/ and https://www.zendesk.com/blog/reports/analyst-reports/total-economic-impact-of-zendesk/ | Contact-rate reduction, AI-agent automated resolution, AHT reduction, agent-workspace productivity, and attrition impact in a vendor-commissioned TEI. | A3, A8 | Planning range / supporting evidence | Medium | Verified vendor-commissioned | Use as support for possible contact-deflection and agent-productivity ranges, not as a universal benchmark. Disclose that it is Zendesk-commissioned and competitor-specific. |
| `SRC-OPS-009` | Ringly / reported Gartner-style benchmark | *Customer Service Cost Statistics 2026* | 2026 | https://www.ringly.io/blog/customer-service-cost-statistics-2026 | Self-service versus assisted-contact cost differential; reported benchmark includes $1.84 self-service versus $13.50 assisted-support channel cost. | A3, A8 | Formula fallback / planning range | Medium | Verified supporting / secondary | Secondary source reporting a Gartner-style benchmark. Prefer customer-provided cost-per-resolution. If used externally, disclose source type and do not present as original Gartner citation unless the original Gartner source is separately obtained. |

### 15.2.4 Loss prevention, shrink, fraud, returns, and safety sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-LP-001` | NRF / LPRC | *2023 National Retail Security Survey* | 2023 | https://nrf.com/research/national-retail-security-survey-2023 | Average shrink 1.6% of sales; $112.1B shrink; segment variance; theft / internal / process attribution. | A6 | Formula input / hero benchmark | High | Verified primary | Strong baseline source for retail shrink; use current customer data when available. NRF discontinued the National Retail Security Survey after the 2023 edition; treat this as the most recent NRF/LPRC baseline and supplement with Appriss Retail or customer-provided shrink data where available. |
| `SRC-LP-001B` | NRF PDF | *2023 National Retail Security Survey* | 2023 | https://a-us.storyblok.com/f/1021220/x/ce11a3e264/nrf_-national_retail_security_survey_2023.pdf | Full PDF for NRF shrink / security findings. | A6 | Formula input / supporting evidence | High | Verified primary | Use when detailed PDF citation is needed. |
| `SRC-LP-002` | Appriss Retail | *The 2026 Total Retail Loss Benchmark Report* | 2026 | https://apprissretail.com/2026-total-retail-loss-benchmark-report/ and https://www.businesswire.com/news/home/20260219293604/en/Appriss-Retail-Benchmark-Report-Uncovers-%24796B-in-Total-Retail-Loss | Total retail loss / preventable loss framing; $796B total retail loss context; returns and shrink pools; preventable loss / preventable shrink framing. | A6, C risk context | Formula input / planning range with ceiling caveat | High | Verified primary / industry report | v1.8 source audit verified the Appriss 2026 TRL source family. Use to support addressable-loss framing, but treat 73% addressable/preventable shrink as a ceiling requiring justification, not a default capture rate. |
| `SRC-LP-003` | Appriss Retail / Deloitte | *2024 Consumer Returns in the Retail Industry* / return and claims fraud analysis | 2024 | https://apprissretail.com/blog/how-return-and-claims-fraud-in-2024-is-shaping-retail-strategies-for-2025/ | Return rate 13.21%; return fraud / abuse 15.14%; $103B return fraud / abuse; empty-box and BORIS/BORO claims-fraud context. | A6 | Planning range / supporting evidence | High | Verified supporting | Use for returns module; distinguish return rate from fraud / abuse share. |
| `SRC-LP-004` | ECR Retail Loss Group | *Self-Checkout Loss in Retail: Global Study* | 2022 | https://ecrloss.com/research-paper/self-checkout-retail-loss-global-study/ | SCO accounts for large share of unknown loss; fixed SCO loss ratio; Scan & Go / BYOD basis-point exposure. | A6 | Formula input / planning range | High | Verified primary | Use for optional SCO submodule when SCO sales mix is known. |
| `SRC-LP-005` | Forrester Consulting / Everseen | *The Total Economic Impact of Everseen Evercheck* | 2023 | https://tei.forrester.com/go/everseen/evercheck/docs/TEI_of_Evercheck.pdf | 0.14% revenue loss reduction; $88K per store; 15% SCO staff productivity improvement; composite TEI. | A6 | Validation only / hero benchmark with disclosure | Medium | Verified vendor-commissioned | Vendor-commissioned TEI; use to validate plausibility, not to anchor default SCO or shrink assumptions. |
| `SRC-LP-006` | Solink | *Solink ROI Study* / market-share study | 2025 | https://solink.com/resources/solink-taking-significant-market-share/ | Investigation-time reduction, theft incident reduction, false-alarm reduction, monthly savings per location. | A6 | Supporting evidence / planning range | Medium | Verified supporting | Vendor / commissioned study; use for range support and disclose source type. |
| `SRC-LP-007` | Spot AI | *From 300 Hours of Footage to the 3 Minutes That Matter* | 2025 | https://www.spot.ai/blog/ai-video-search-retail | AI video search time reduction; investigation compression; case-study shrink reduction examples. | A6 | Supporting evidence / upside case | Medium | Verified supporting | Vendor case study; do not treat individual retailer results as universal. |
| `SRC-LP-008` | Appriss Retail | *The Curious Case of Phantom Inventory* | 2024 | https://apprissretail.com/blog/the-curious-case-of-phantom-inventory/ | Phantom inventory, on-shelf availability, stockout / lost-sales connection. | A6, A2 | Supporting evidence / optional upside | Medium | Verified supporting | Use for inventory accuracy / OSA discussion; avoid stacking with shrink unless separated. |
| `SRC-LP-009` | RELEX Solutions | *How to Optimize Cycle Counting and Banish Phantom Inventory* | 2024 | https://www.relexsolutions.com/resources/how-to-optimize-cycle-counting-and-banish-phantom-inventory/ | Inventory record inaccuracy and phantom inventory cost context. | A6 | Supporting evidence | Medium | Verified supporting | Vendor source; use as context, not formula-grade proof. |
| `SRC-LP-010` | NRF / LPRC | *The Impact of Retail Theft & Violence 2025* | 2025 | https://nrf.com/media-center/press-releases/new-study-finds-retailers-continue-to-contend-with-rising-levels-of-theft-and-violence | Shoplifting incident increase, violence increase, ORC involvement, underreporting context. | A6, C risk context | Supporting evidence / protected-value support | High | Verified primary | Use for safety and risk context, not direct cash savings by default. |
| `SRC-LP-011` | NCCI | *Workplace Violence: Part One - Assault Trends, Drivers, and Demographics* | 2026 | https://www.ncci.com/Articles/Pages/Insights_Workplace-Violence-Part1-Assault-Trends-Drivers-Demographics.aspx | Nonfatal assault trend; retail trade as a major sector for assaults. | A6, C risk context | Protected-value support | High | Verified primary | Use for workplace violence trend support; not retail-specific cost formula by itself. |
| `SRC-LP-012` | TimeForge | *Workplace Violence Prevention Programs Reduce Safety Risks* | 2024 | https://timeforge.com/industry-news/prevention-programs-address-workplace-violence-risks/ | Safety training / prevention, turnover and insurance-premium context. | C4 risk context | Supporting evidence only | Directional | Manual verification required | Industry article; do not use as formula input. Use only with caveats and customer-specific turnover / insurance data. |

### 15.2.5 IT resilience, observability, downtime, Cisco infrastructure, and connected-store technology sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-IT-001` | Gartner legacy downtime benchmark; Splunk / Oxford Economics; New Relic-style retail outage context | Downtime-cost benchmark family | 2014 / 2024 / 2025 | https://www.splunk.com/en_us/form/the-hidden-costs-of-downtime.html | Legacy $5,600/minute-style downtime benchmark; newer cross-industry and retail / ecommerce outage benchmarks used only as tiered Layer C context. | A13, C1 | Hero benchmark / supporting evidence; not a formula default | Medium-High | Verified supporting / context | Keep legacy Gartner-style figure only as conservative historical lower-bound context. Do not use any downtime benchmark as a hard default; customer-provided downtime exposure should drive calculations. Some source pages may be gated or secondary. |
| `SRC-IT-002` | ThousandEyes / Cisco | *Retail Digital Experience Monitoring* | 2025 | https://www.thousandeyes.com/solutions/retail | Retail digital experience monitoring, cloud / SaaS / network dependency visibility. | A11, A12, A13, C1 | Cisco relevance / supporting evidence | High | Verified supporting | Cisco-owned. Use for platform / observability mapping, not direct value benchmark. |
| `SRC-IT-003` | ThousandEyes / Cisco | *Enhancing Visibility With the ThousandEyes and Meraki Integration* | 2025 | https://www.thousandeyes.com/blog/enhancing-visibility-thousandeyes-meraki-integration | ThousandEyes + Meraki visibility and troubleshooting. | A11, A12, A13 | Cisco relevance / supporting evidence | High | Verified supporting | Cisco-owned. Useful for portfolio fit. |
| `SRC-IT-004` | Cisco Meraki Documentation | *Meraki MX ThousandEyes Configuration Guide* | 2025 | https://documentation.meraki.com/SASE_and_SD-WAN/MX/Integrations/MI_-_Meraki_Insight/Product_Information_and_Configuration/Meraki_MX_ThousandEyes_Configuration_Guide | Technical proof of Meraki / ThousandEyes integration. | A11, A12, A13 | Cisco relevance | High | Verified supporting | Cisco-owned technical documentation. |
| `SRC-TEV-005` | Cisco / Forrester Consulting | *The Total Economic Impact of Cisco Full-Stack Observability* | 2026 | https://tei.forrester.com/go/cisco/FSO//docs/TheTEIOfCiscoFull-StackObservability.pdf | Full-stack observability value, incident reduction, MTTR, operational visibility. | A13, C1 | Formula input / Cisco proof point | High | Verified vendor-commissioned | Cisco-commissioned TEI; disclose. |
| `SRC-TEV-006` | Cisco / Forrester Consulting | *Cisco Meraki Economic Impact Study* | 2026 | https://www.cisco.com/site/us/en/products/networking/networking-cloud/forrester-tei-study-cloud-platform.html | Meraki cloud platform economic impact, network operations, remote management. | A12, A11, A10 | Formula input / Cisco proof point | High | Verified vendor-commissioned | Cisco-commissioned TEI; disclose. |
| `SRC-CSCO-001` | Cisco Spaces | *Cisco Spaces for Retail* | 2026 | https://www.cisco.com/c/en/us/solutions/enterprise-networks/spaces/retail.html | Cisco Spaces / location intelligence, footfall, behavior, and store analytics relevance. | A1, A7, B1 | Cisco relevance | High | Verified supporting | Use for Cisco capability fit, not external value benchmark. |
| `SRC-CSCO-002` | Cisco on Cisco | *Cisco Store Retail Analytics* / *How new data streams transformed Cisco Store decision-making* | 2025 / 2026 | https://www.cisco.com/site/uk/en/solutions/cisco-on-cisco/cisco-store-retail-analytics.html and https://blogs.cisco.com/cisco-on-cisco/how-new-data-streams-transformed-cisco-stores-decision-making | Cisco Store retail analytics, dwell, energy, telemetry, and operational proof points. | A7, A10, B1 | Cisco proof point | High | Verified supporting | Cisco-owned case study; do not generalize as universal benchmark. |
| `SRC-CSCO-003` | Cisco / Webex | *Cisco Store / Webex Connect Retail Case Study* | 2025 | https://www.webex.com/content/dam/www/us/en/documents/customers/case-study/us-cisco-store-case-study-july2025.pdf | Webex Connect retail journey / CPaaS proof point. | A3, A8, B3 | Cisco proof point | High | Verified supporting | Cisco-owned case study; relevant to CPaaS, not a universal benchmark. |
| `SRC-CSCO-004` | Cisco Meraki / Forrester | *Cisco Meraki Economic Impact Study* | 2026 | https://tei.forrester.com/go/cisco/merakinetworking/ | Meraki economic impact / operating efficiency. | A10, A12 | Cisco proof point / formula support | High | Verified vendor-commissioned | Cisco-commissioned TEI; overlaps with `SRC-TEV-006`. |
| `SRC-CSCO-005` | Cisco security portfolio | Duo, ISE, Secure Access / SASE, XDR product pages | 2026 | https://duo.com/product/multi-factor-authentication-mfa ; https://www.cisco.com/c/m/en_uk/products/security/identity-services-engine/use-case-segmentation.html ; https://www.cisco.com/site/us/en/solutions/secure-access-service-edge-sase/index.html ; https://www.cisco.com/site/us/en/products/security/xdr/index.html | Cisco identity, segmentation, secure access, and XDR capability fit. | C2, C3 | Cisco relevance | High | Verified supporting | Replaces generic Cisco homepage / `SRC-CYB-008` issue. |

### 15.2.6 Cybersecurity, zero trust, segmentation, and data breach sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-CYB-001` | IBM | *Cost of a Data Breach Report 2025* | 2025 | https://www.ibm.com/reports/data-breach | Breach cost, breach lifecycle, security AI / automation, US / global cost context. | C2, C3 | Formula input / protected-value benchmark | High | Verified primary | Use as primary cyber exposure benchmark. Do not let user-entered cyber exposure produce unrealistic values without caveat. |
| `SRC-CYB-002` | Verizon | *2025 Data Breach Investigations Report* | 2025 | https://www.verizon.com/business/resources/T83c/reports/2025-dbir-data-breach-investigations-report.pdf | Breach vectors, credential risk, ransomware, vulnerability exploitation, third-party risk. | C2, C3 | Formula input / supporting evidence | High | Verified primary | Strong source for threat patterns. |
| `SRC-CYB-003` | Sophos | *The State of Ransomware in Retail 2024* | 2024 | https://www.sophos.com/en-us/blog/the-state-of-ransomware-in-retail-2024 | Retail ransomware disruption, recovery, impact. | C2 | Supporting evidence / hero benchmark | High | Verified supporting | Security vendor research; retail-specific. |
| `SRC-CYB-004` | Cisco Duo | *Duo MFA: Phishing-Resistant Identity Protection* | 2026 | https://duo.com/product/multi-factor-authentication-mfa | MFA, identity protection, secure access. | C2, C3 | Cisco relevance | High | Verified supporting | Capability support, not external benchmark. |
| `SRC-CYB-005` | Cisco ISE | *Cisco ISE segmentation use case* | 2026 | https://www.cisco.com/c/m/en_uk/products/security/identity-services-engine/use-case-segmentation.html | Identity Services Engine / segmentation capability context. | C3 | Cisco relevance | High | Verified supporting | Capability support. |
| `SRC-CYB-006` | Cisco Secure Access / SASE | *Secure Access Service Edge / SASE* | 2026 | https://www.cisco.com/site/us/en/solutions/secure-access-service-edge-sase/index.html | Secure access / SASE / zero-trust capability context. | C2, C3, A12 | Cisco relevance | High | Verified supporting | Capability support. |
| `SRC-CYB-007` | Cisco XDR | *Cisco XDR - Extended Detection and Response* | 2026 | https://www.cisco.com/site/us/en/products/security/xdr/index.html | Detection and response / telemetry correlation capability context. | C2, C3 | Cisco relevance | High | Verified supporting | Capability support. |
| `SRC-CYB-008` | Retired generic Cisco homepage source | Generic Cisco homepage | 2026 | N/A | Previously generic Cisco security citation. | C2, C3 | Retired | N/A | Remove | Do not use. Replaced by `SRC-CSCO-005`, `SRC-CYB-004`, `SRC-CYB-005`, `SRC-CYB-006`, and `SRC-CYB-007`. |

### 15.2.7 AI-native retail, store strategy, and future-option sources

| Source ID | Organization | Source title | Year | URL | Benchmarks / claims used | Applies to | Model role | Confidence | Status | Notes / caveats |
|---|---|---|---:|---|---|---|---|---|---|---|
| `SRC-AI-001` | McKinsey / ICSC | *Shopping in the Age of AI: Redefining Stores for a New Era* | 2026 | https://www.mckinsey.com/industries/retail/our-insights/shopping-in-the-age-of-ai-redefining-stores-for-a-new-era | AI-native store macro thesis, changing store role, customer expectations. | A1, A2, D | Supporting evidence | High | Manual verification required | May be bot-blocked in automated checks. Strong strategic source if manually accessible. |
| `SRC-AI-002` | Deloitte | *2025 US Retail Industry Outlook* | 2025 | https://www.deloitte.com/us/en/insights/industry/retail-distribution/retail-distribution-industry-outlook/2025.html | Retail AI, personalization, operating outlook. | A1, A2, D | Supporting evidence | High | Verified primary | Strong strategic retail source. |
| `SRC-AI-003` | Capgemini Research Institute | *What Matters to Today’s Consumer 2025* | 2025 | https://www.capgemini.com/insights/research-library/what-matters-to-todays-consumer-2025/ | Consumer expectations, AI shopping behavior, personalization context. | A1, D | Supporting evidence | High | Verified primary | Strong consumer / retail context. |
| `SRC-AI-004` | BCG | *Retail Rewired: How AI Is Reshaping the Retail Business Model* | 2026 | https://www.bcg.com/publications/2026/how-ai-is-reshaping-the-retail-business-model | AI reshaping retail business model and capabilities. | D, A1-A13 context | Supporting evidence | High | Manual verification required | May be bot-blocked; manually verify before external distribution. |
| `SRC-TEV-007` | McKinsey / QuantumBlack | *The Agentic Commerce Opportunity* | 2026 | https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-agentic-commerce-opportunity | Agentic commerce future-option context. | D | Supporting evidence / hero context | High | Manual verification required | Use for Layer D future option only, not Layer A. |

---

## 15.3 Citation expectations by value driver

When an agent generates a TEV output, it should include a short “source support” section that names the source IDs used for each modeled value driver. The agent does not need to list every URL in the executive summary, but it must be able to provide the full source registry when asked.

Example output note:

```text
Source support used:
- A3 AI-Powered Service and Outreach: SRC-FPD-003, SRC-CSCO-003, SRC-OPS-008, SRC-OPS-009.
- A4 Dynamic Digital Merchandising and Retail Media: SRC-RM-001, SRC-RM-002, SRC-RM-003, SRC-RM-006.
- B3 Consented Messaging Identities: SRC-FPD-003, SRC-FPD-007, SRC-FPD-008.
```

For Cisco review, always preserve the source IDs, organization names, source titles, years, URLs, source status, model role, and caveats in the Markdown file.

---

## 15.4 Source-use guardrails

1. **Use original sources only.** Do not cite internal research compilation documents as the benchmark source in customer-facing output.
2. **Distinguish source type.** Identify whether a source is analyst, standards body, retailer proof point, Cisco proof point, vendor benchmark, market research, or directional evidence.
3. **Disclose vendor-commissioned TEIs.** Forrester TEI and similar analyst studies commissioned by vendors are useful modeled case-study evidence, but they must be labeled as vendor-commissioned and should not be presented as universal independent benchmarks.
4. **Do not overclaim vendor benchmarks.** Vendor sources can support ranges and examples, but customer data should replace them where possible.
5. **Do not turn hero benchmarks into formula assumptions automatically.** Hero benchmarks should inspire ranges; formulas should use conservative/base/upside assumptions.
6. **Preserve Layer B framing.** Sources supporting first-party data value, messaging identity value, retail media addressability, and product-intent signals support a strategic asset proxy, not booked revenue.
7. **Preserve Layer C framing.** Cyber and downtime sources support protected value at risk, not new revenue.
8. **Use customer-specific numbers first.** If a retailer provides actual CAC, CPM, RPS, shrink, contact cost, outage cost, or ESL labor data, use that instead of default benchmarks.
9. **Flag source confidence and status.** If a value driver relies primarily on medium-confidence vendor sources, manual-verification-required sources, or retired / replacement sources, label the output as benchmark-informed and directional.
10. **Do not use retired benchmarks.** Do not use `SRC-CYB-008`, `SRC-RM-007`, `SRC-OPS-006`, `SRC-OPS-007`, or the retired `$3.10 back-in-stock RPS` claim unless they are replaced and reverified.
11. **Treat LP sources by role.** NRF/LPRC, Appriss, and ECR Retail Loss can support formula inputs or planning ranges; Forrester TEI and vendor case studies should validate ranges or support upside scenarios unless customer data confirms comparable conditions.
12. **Use cost-per-resolution for service economics when available.** Contact automation can change the mix of remaining live-agent contacts. If customer data is available, model successfully resolved routine inquiries and cost per resolution instead of treating all avoided contacts as equal.
13. **Avoid source creep.** New inventory, RFID, downtime, and GenAI sources should strengthen citations or optional scenario support; they should not automatically create new default formulas.

---

## 15.5 Quick source lookup by model component

| Model component | Preferred source IDs |
|---|---|
| Enriched CRM profile value | `SRC-FPD-001`, `SRC-FPD-002`, `SRC-FPD-011`, `SRC-FPD-012`, `SRC-FPD-013` |
| Known store session value | `SRC-FPD-004`, `SRC-FPD-005`, `SRC-FPD-006`, `SRC-AI-001`, `SRC-CSCO-001` |
| Consented messaging identity value | `SRC-FPD-003`, `SRC-FPD-007`, `SRC-FPD-008`, `SRC-CSCO-003` |
| Product-intent / zero-party value | `SRC-FPD-009`, `SRC-FPD-010`, `SRC-FPD-014`, `SRC-SHELF-001`, `SRC-SHELF-002` |
| Retail media CPM / media value | `SRC-RM-001`, `SRC-RM-002`, `SRC-RM-003`, `SRC-RM-004`, `SRC-RM-006` |
| Queue and service-flow value | `SRC-OPS-001`, `SRC-OPS-002`, `SRC-CSCO-001`, `SRC-CSCO-002` |
| AI support / contact deflection | `SRC-FPD-003`, `SRC-CSCO-003`, `SRC-OPS-008`, `SRC-OPS-009` |
| ESL / price-change labor | `SRC-OPS-003`, `SRC-OPS-004`, `SRC-OPS-005`, `SRC-SHELF-001` |
| Shrink / loss prevention | `SRC-LP-001`, `SRC-LP-001B`, `SRC-LP-002`, `SRC-LP-003`, `SRC-LP-004`, `SRC-LP-005`, `SRC-LP-006`, `SRC-LP-007`, `SRC-LP-008`, `SRC-LP-009`, `SRC-LP-010`, `SRC-LP-011`, `SRC-LP-012` |
| Inventory accuracy / OSA optional scenario | `SRC-INV-001`, `SRC-INV-002`, `SRC-INV-003`, `SRC-INV-004`, `SRC-LP-008`, `SRC-LP-009` |
| Smart lighting / facilities | `SRC-CSCO-002`, `SRC-CSCO-004`, `SRC-TEV-006` |
| Device reliability / endpoint visibility | `SRC-OPS-001`, `SRC-IT-002`, `SRC-IT-003`, `SRC-IT-004`, `SRC-TEV-006` |
| Observability / MTTR | `SRC-TEV-005`, `SRC-IT-001`, `SRC-IT-002`, `SRC-IT-003`, `SRC-IT-004` |
| Cyber / zero trust | `SRC-CYB-001`, `SRC-CYB-002`, `SRC-CYB-003`, `SRC-CYB-004`, `SRC-CYB-005`, `SRC-CYB-006`, `SRC-CYB-007`, `SRC-CSCO-005` |
| AI future option value | `SRC-AI-001`, `SRC-AI-002`, `SRC-AI-003`, `SRC-AI-004`, `SRC-TEV-007` |

---

## 15.6 Missing, blocked, or unverifiable source treatment

If a conversational agent cannot access a linked source, it should still name the source ID and disclose that the source link should be manually verified during Cisco review. Do not invent a statistic or claim that a source says something beyond what is listed in this registry.

If a source is bot-blocked but likely valid, say:

```text
This source is listed in the TEV registry but may require manual browser verification because automated checks can be blocked by the publisher.
```

If a claim is not found in the linked source, say:

```text
The source supports the general concept, but the exact benchmark claim has not been verified. Treat this as directional until replaced with a verified primary source.
```

---

## 15.7 Layer B triangulation methodology

Layer B proxy values are **not copied from one source that says “an enriched CRM profile is worth $X.”** They are triangulated from several economic evidence families:

1. **Replacement-cost evidence:** Paid acquisition, CPL, CPA, CPI, and retained-user acquisition benchmarks support the avoided-cost value of acquiring known customers organically through Wi-Fi, app, loyalty, messaging, QR / GS1 2D, and in-store interactions. Relevant source IDs include `SRC-FPD-011` and `SRC-FPD-012`.
2. **Activation-value evidence:** CDP, profile resolution, personalization, and messaging activation sources show that known, unified profiles can improve conversion, messaging effectiveness, media suppression, and marketing productivity. Relevant source IDs include `SRC-FPD-001`, `SRC-FPD-002`, `SRC-FPD-003`, and `SRC-FPD-013`.
3. **Messaging-identity evidence:** Conversational commerce, SMS / RCS performance, and Webex Connect retail proof points support the idea that a consented phone / messaging identity is more valuable than a generic anonymous session. Relevant source IDs include `SRC-FPD-003`, `SRC-FPD-007`, `SRC-FPD-008`, and `SRC-CSCO-003`.
4. **Loyalty and retention evidence:** Loyalty / retention sources support the strategic value of converting anonymous store behavior into known customer relationships and improving engagement over time. Relevant source IDs include `SRC-FPD-004`, `SRC-FPD-005`, and `SRC-FPD-006`.
5. **Retail media addressability evidence:** In-store retail media standards and retailer proof points support the value of deterministic audiences, addressable impressions, and closed-loop measurement. Relevant source IDs include `SRC-RM-001`, `SRC-RM-002`, `SRC-RM-003`, `SRC-RM-004`, and `SRC-RM-006`.
6. **Product-intent / zero-party evidence:** Quiz, preference, QR, GS1 2D, and shelf-edge interaction sources support the strategic value of declared intent and product-specific questions. Relevant source IDs include `SRC-FPD-009`, `SRC-FPD-010`, `SRC-FPD-014`, `SRC-SHELF-001`, and `SRC-SHELF-002`.

Therefore, when explaining Layer B, the agent should use language like:

```text
Layer B values are triangulated strategic asset proxies based on replacement cost, activation value, messaging identity economics, loyalty economics, retail media addressability, and product-intent signal value. They are not annual booked revenue or profit.
```

Do not say that any single source proves the entire Layer B value range. Cite the source families used to support the relevant proxy component.

---

## 15.8 Retired, replaced, or demoted sources and benchmark claims

| Source / claim | v1.4 / v1.5 treatment | Reason |
|---|---|---|
| G2-hosted Amperity TEI PDF | Replaced by official Amperity TEI page (`SRC-FPD-001`). | Third-party file host / source-stability risk. |
| LoyaltyPoint retention-profit citation | Replaced by HBR / Bain Reichheld-Sasser sources (`SRC-FPD-004`, `SRC-FPD-005`). | Famous stat is commonly misattributed; primary source is stronger. |
| Generic `https://www.cisco.com/` security citation | Retired (`SRC-CYB-008`); replaced by Duo, ISE, Secure Access, XDR, and Cisco security pages. | Generic homepage is not a citation. |
| Osmos in-store media link as primary support | Retired as primary (`SRC-RM-007`). | Audit reported connection refused / unstable link. |
| `$3.10 back-in-stock RPS` claim | Retired until reverified. | Audit reported exact claim not found in cited source. |
| MRI Software / ScanQueue queue sources | Retired as primary support. | Red-team flagged as wrong-domain / obscure for queue data. |
| Former `SRC-RM-005` registry slot | Intentionally retired / skipped. | Earlier retail-media numbering changed during source cleanup. Use the active retail-media IDs listed in Section 15.5; do not cite `SRC-RM-005`. |
| Former manual LP placeholder `SRC-LP-002` | Replaced in v1.5 by Appriss Retail 2026 Total Retail Loss Benchmark and promoted in v1.8 based on the source audit. | Earlier entry had N/A URL. Current entry provides Appriss / Business Wire source URLs and role as addressable / preventable loss-pool support, with the 73% preventable-shrink figure still treated as a ceiling. |
| Former manual LP placeholder `SRC-LP-003` | Replaced in v1.5 by ECR Retail Loss, Forrester/Everseen, Solink, and Spot AI sources. | Earlier entry had N/A URL and combined too many concepts. New entries separate SCO exposure, TEI validation, investigation productivity, and vendor case-study support. |
| Former manual LP placeholder `SRC-LP-004` | Replaced in v1.5 by NRF/LPRC, NCCI, and TimeForge safety sources. | Earlier entry had N/A URL. New entries distinguish retail violence context, workers' compensation / assault trend context, and safety / turnover support. |
| Healthcare workplace violence benchmarks | Avoid as retail formula input. | Healthcare safety case studies may be useful analogies but are not retail-specific and should not drive A6 formula defaults. |
| Claims that every detected fraud event converts into recovered cash | Avoid. | Detected exceptions, prevented incidents, and recovered value are different concepts. Use conservative reduction assumptions and disclose the difference. |
| Former `SRC-INFRA-001` registry reference | Replaced by `SRC-CSCO-002`. | Cisco Store / Cisco-on-Cisco facilities and analytics support now resolves under Cisco-specific source IDs. |
| Former `SRC-TEV-004` registry reference | Replaced by `SRC-OPS-008`. | Zendesk / Forrester TEI customer-service automation support now resolves under operations source IDs. |
| Pairwise overlap haircut ranges | Labeled as TEV methodology convention, not external benchmark. | These are model-governance controls, not third-party statistics. |

---

## 15.9 Source requirements for Cisco-facing output

For any TEV output intended for Cisco review or customer sharing, the agent should include:

- [ ] The value drivers used.
- [ ] The source IDs supporting each major value driver.
- [ ] Confirmation that every cited source ID resolves to an active registry entry or to a retired / replacement note in Section 15.8.
- [ ] Whether each source is primary, vendor-commissioned, supporting, or manual-verification-required.
- [ ] Any retired / replacement-source warnings.
- [ ] A statement that customer-specific data should replace benchmark defaults wherever available.



# 16. Profit / ROI / TCO Conversation Guardrail

This TEV calculator should not calculate profit, ROI, payback, or TCO by default.

However, a user may ask follow-up questions such as:

- “What portion of this could hit gross profit?”
- “What does this look like after margin?”
- “Can you estimate EBITDA contribution?”
- “What would finance care about?”
- “What is actual savings versus influenced value?”
- “What would ROI look like if the program cost is $X?”
- “What is the TCO?”

If that happens, explain that this requires a separate finance sensitivity discussion. Do not alter the core TEV model.

## 16.1 TEV drivers that could support a finance sensitivity discussion

| TEV driver type | Possible finance sensitivity |
|---|---|
| Recovered / influenced transaction value | Could be margin-adjusted using customer-provided gross, contribution, operating, or EBITDA margin. |
| Qualified commercial actions | Could be margin-adjusted if the user asks for profit contribution. |
| Avoided contact cost | Already close to avoided cost; may need realization factor. |
| Labor time saved | May need realization factor or proof of payroll/overtime reduction. |
| Energy savings | Usually direct avoided utility cost. |
| Shrink reduction | Usually recovered value / margin protection; finance may decide treatment. |
| Retail media economics | Could be adjusted using media contribution economics. |
| Layer B strategic proxy | Usually not profit; keep separate unless activated and measured. |
| Layer C protected value | Not new revenue; keep non-additive. |
| Layer D option value | Not current-period value unless activated and measured. |

## 16.2 Inputs required for a separate finance sensitivity

Only ask for these if the user explicitly requests profit, ROI, payback, or TCO:

- gross margin;
- contribution margin;
- operating margin or EBITDA margin;
- Cisco / program cost;
- implementation cost;
- annual run cost;
- deployment timeline;
- expected realization rate;
- whether labor savings become actual payroll reduction, overtime reduction, or reallocated capacity.

## 16.3 Required label

Any such analysis must be shown as a **range**, not a single-point profit or ROI estimate, and must be labeled:

```text
Optional finance sensitivity, not the core TEV model.
```

The label must appear directly above the sensitivity output. Do not allow the sensitivity to replace the core TEV view.

## 16.4 What not to do

Do not:

- convert every revenue-related TEV driver to margin by default;
- call Layer A profit;
- include Layer B in ROI;
- include Layer C in ROI;
- treat protected value as new revenue;
- treat future option value as current-year cash flow;
- calculate payback unless program cost is provided;
- present any sensitivity as guaranteed;
- present a single-point profit, ROI, or payback estimate without a range and disclaimer.

---

# 17. Output language examples

## Good language

- “The model estimates $X in Layer A direct annual TEV impact.”
- “Layer B represents a strategic proxy for first-party context value, not booked revenue.”
- “Layer C is protected value at risk and should not be added to Layer A as new revenue.”
- “This estimate is benchmark-informed and should be refined with customer-specific data.”
- “The largest assumptions are identified-session rate, shrink rate, contact volume, and deployment scope.”

## Avoid language

- “Cisco will generate $X in profit.”
- “This guarantees an ROI of X%.”
- “Layer B is revenue.”
- “Protected value should be added to the annual return.”
- “Every scan converts to a purchase.”
- “Every labor minute saved reduces payroll.”

---

# 18. Worked example instructions

When a user gives a scenario such as:

```text
Retailer with $80B annual revenue and 2,000 stores
```

Do the following:

1. Confirm known inputs.
2. Ask whether they know average basket, visits, shrink rate, contact volume, energy spend, and IT incident volume.
3. If they want a quick estimate, use conservative defaults and clearly label the output as directional.
4. Build conservative/base/upside cases.
5. Show Layer A, B, C, and D separately.
6. Highlight top value drivers.
7. Provide next discovery questions.

Do not assume the retailer has mature identity, retail media, ESL, smart lighting, or observability capabilities unless the user says so.

## 18.1 Worked LP example instructions

When A6 is active, show how caps and overlap controls were applied. Use a compact worked example when SCO, returns, or phantom inventory are included.

Example format:

```text
A6 LP example:
Annual revenue: $5B
Shrink rate: 1.6%
Total shrink pool: $80M
Addressable shrink: 50% = $40M
Reduction of addressable shrink: 15%
Shrink reduction value: $6M

SCO module: enabled because customer provided SCO mix.
SCO sales mix: 30%
SCO basis point assumption: 2.6 bps
SCO exposure: $5B × 30% × (2.6 / 10,000) = $390K

A6 umbrella check:
Customer-confirmed / benchmark-derived addressable loss pool: $40M
A6 total before overlap: $X
A6 cap at 60-70% of addressable loss pool: $24M-$28M
A6 output after overlap/cap: $Y
```

If the customer did not provide SCO mix or OSA baseline, state that those modules remain off.

---

# 19. Final calculator behavior checklist

Before giving any TEV output, verify:

- [ ] Layer A, B, C, and D are separated.
- [ ] Layer B is labeled as strategic asset value, not annual cash flow, booked revenue, ROI, or profit.
- [ ] Layer B is shown as a range, with the strategic asset-value disclaimer adjacent to the number.
- [ ] If Layer B exceeds 25% of annual revenue, the output flags it as a large strategic asset proxy and does not present it as revenue.
- [ ] Layer C is not added as direct annual impact.
- [ ] Layer D is not counted as current value unless activated.
- [ ] Gross TEV drivers are not called profit.
- [ ] ROI/TCO is not calculated by default.
- [ ] Overlap haircuts are applied to Layer A.
- [ ] A6 LP submodule overlaps and the A6 umbrella cap are checked when LP is active.
- [ ] SCO and phantom inventory / OSA remain off unless required customer inputs are provided.
- [ ] Safety quantification is kept in Layer C unless customer-specific cost data is provided.
- [ ] Labor realization is handled carefully.
- [ ] Customer-provided inputs and defaults are listed.
- [ ] Confidence level is stated.
- [ ] Source IDs from Section 15 are cited for major value drivers used in the output.
- [ ] Every source ID cited in the output resolves to a registry entry and includes a source-status label.
- [ ] Manual-verification sources are marked with `(MV)` and are not the sole support for formula defaults or headline values.
- [ ] All cited source IDs are resolvable in the Benchmark Source Registry or intentionally retired / replaced in Section 15.8.
- [ ] Vendor-commissioned TEI sources are disclosed when used.
- [ ] Retired or manual-verification-required sources are not presented as fully verified.
- [ ] Next discovery questions are included.

---

# 20. Short version for agent instructions

Use this summary if the conversational agent needs quick guidance:

```text
You are calculating Cisco Retail 2.0 Total Economic Value, not ROI or TCO.
Model at the value-driver level.
Separate outputs into Layer A direct annual impact, Layer B First-Party Context Engine strategic proxy, Layer C protected value at risk, and Layer D future option value.
Do not margin-adjust by default.
Do not calculate profit, ROI, payback, or TCO unless explicitly asked, and then label it as a separate finance sensitivity.
Use customer inputs first and conservative defaults second.
Apply overlap controls.
Do not double-count the same transaction, contact, labor hour, shrink dollar, incident, profile, or protected value pool.
Always explain assumptions, confidence, source IDs, and source-status caveats.
```


---

# 7. TEV Input Intake Template

# Cisco Retail 2.0 TEV Input Intake Template

## Purpose

This intake template is designed to collect the minimum practical inputs needed to run the Cisco Retail 2.0 Total Economic Value (TEV) Calculator.

The TEV calculator is an early-stage scenario tool. It is not an ROI, TCO, NPV, payback, profit, or implementation-cost model. Unknown inputs are acceptable. When a customer does not know an input, the calculator should use conservative benchmark defaults and clearly label them.

---

## How to use this template

Use this template before running the TEV calculator. The goal is not to collect every data point perfectly. The goal is to gather enough customer context to create a credible first-pass TEV scenario, identify the highest-value areas, and guide the next discovery conversation.

For each field, mark one of the following:

- **Provided**: customer gave the value
- **Estimate**: customer gave an approximate value
- **Unknown**: use conservative benchmark default
- **Not applicable**: driver does not apply to this retailer

---

# 1. Customer and scenario context

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Retailer / account name |  |  |  |
| Retail segment |  |  | Grocery, pharmacy, mass merchant, apparel, convenience, specialty, department store, other |
| Geography / market |  |  | U.S., North America, global, etc. |
| Scenario owner |  |  | Cisco rep / partner / customer contact |
| Scenario date |  |  |  |
| Scenario purpose |  |  | Discovery, executive readout, value workshop, partner planning, etc. |
| Deployment scope |  |  | Enterprise-wide, region, pilot stores, priority format, etc. |
| Number of stores in scope |  |  | If different from total store count |
| Timeframe represented | Steady-state annual TEV | Provided | The model should default to steady-state annual TEV, not Year 1 ramp |

---

# 2. Company scale inputs

These inputs are used across multiple value drivers.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Annual revenue |  |  | Total revenue or retail revenue in scope |
| Store count |  |  | Total physical stores |
| Annual store visits |  |  | Use customer data if available; otherwise estimate |
| Annual transactions |  |  | Optional alternative to visits |
| Average basket / transaction value |  |  | Used for commercial-action and recovered transaction value |
| Annual ecommerce / omnichannel revenue |  |  | Optional; useful for BOPIS, returns, and journey orchestration |
| Annual BOPIS / curbside / pickup orders |  |  | Optional |
| Annual return value |  |  | Gross dollar value of returned merchandise at retail price |
| Deployment percentage |  |  | % of stores / workflows included in scenario |

**Important:** Do not ask for gross margin, operating margin, EBITDA margin, or program cost for the core TEV model. Those belong only in a separate finance sensitivity conversation if explicitly requested.

---

# 3. Customer identity and first-party context inputs

These inputs support Layer B: First-Party Context Engine strategic asset value proxy.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Loyalty members |  |  | Total enrolled loyalty members |
| Active loyalty members |  |  | If known |
| Active app users |  |  | Monthly or annual active users |
| Email-reachable customers |  |  | CRM / email database size |
| SMS / RCS / WhatsApp / phone-reachable customers |  |  | Consented messaging identities |
| Identified store session rate |  |  | % of store visits tied to loyalty, app, Wi-Fi, phone, email, or CRM identity |
| Wi-Fi onboarding / captive portal opt-in rate |  |  | If applicable |
| Cisco Spaces / location intelligence in use today? |  |  | Yes / No / Planned |
| Average visits per identified customer |  |  | Optional; helps deduplicate known sessions |
| CRM profile enrichment rate |  |  | % of known profiles receiving useful new context |
| Consent / channel eligibility constraints |  |  | Important for messaging identity valuation |

## Layer B guardrail

Layer B is strategic asset value. It is not booked revenue, annual cash flow, ROI, profit, or TCO savings. If Layer B exceeds 25% of annual revenue, flag it. If it exceeds 50% of annual revenue, require explicit user acknowledgment and recommend conservative/base assumptions.

---

# 4. Shelf-edge scan / QR / GS1 2D / product-intent inputs

These inputs support scan-to-learn, product intent, zero-party data, and optional scan-originated commercial actions.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| SKU count in scope |  |  | Products / categories with shelf-edge codes, QR, GS1 2D, ESL, or paper labels |
| Shelf-edge / QR / GS1 2D scans per year |  |  | If unknown, use conservative benchmark assumption |
| Scan rate |  |  | Scans as % of visits, transactions, or labels |
| Opt-in / match rate from scans |  |  | % tied to email, phone, loyalty, app, or CRM identity |
| Qualified product-intent rate |  |  | % of scans producing useful intent |
| Zero-party declared preference captures |  |  | Allergens, compatibility, recipe interest, preference, coupon intent, etc. |
| Scan-to-commercial-action rate |  |  | Optional Layer A only if measured or customer-provided |
| Average scan-originated action value |  |  | Optional |
| Product Q&A containment rate |  |  | If scan-to-AI service reduces assisted contacts |

## Scan guardrail

A scan can produce multiple value pools: direct action value, product-intent signal value, zero-party preference value, and CRM identity value. Do not double-count the same captured data field twice inside Layer B.

---

# 5. Customer service, messaging, and contact center inputs

These inputs support AI-powered service, outreach, CPaaS, Webex Contact Center, Webex AI, and workflow protection.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Annual customer contacts |  |  | All service contacts across phone, chat, email, messaging, etc. |
| Routine inquiry percentage |  |  | % addressable by automation / AI |
| Current deflection / self-service rate |  |  | If known |
| Target deflection rate |  |  | Conservative/base/upside |
| Cost per assisted contact |  |  | Customer-provided preferred |
| Cost per self-service interaction |  |  | Optional |
| Average handle time |  |  | Minutes |
| Agent cost per minute |  |  | Optional |
| First-contact resolution rate |  |  | Use for cost-per-resolution logic |
| Repeat contact rate |  |  | Optional |
| Messaging channels in use |  |  | SMS, RCS, WhatsApp, app push, email, Webex Connect, etc. |
| Annual outbound messages |  |  | Optional |
| Annual inbound messaging conversations |  |  | Useful for consented messaging identity value |

## Contact-center guardrail

Prefer cost-per-resolution logic over simple cost-per-contact when first-contact resolution or repeat contact data is available. Automation can deflect routine contacts while making the remaining live-agent contacts more complex.

---

# 6. Retail media and digital merchandising inputs

These inputs support dynamic digital merchandising, retail media, in-store screens, signage, and media addressability.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Existing retail media network? |  |  | Yes / No / Planned |
| In-store media monetization active? |  |  | Yes / No / Planned |
| Annual store traffic in media-enabled locations |  |  |  |
| Number of screens / touchpoints in scope |  |  | Screens, ESLs, kiosks, app placements, shelf-edge scans |
| Estimated impressions per visit |  |  | Optional |
| Fill rate |  |  | % of available inventory sold / used |
| CPM assumption |  |  | Use conservative unless customer has active RMN |
| Supplier / brand campaign count |  |  | Optional |
| Category sales in scope |  |  | For category lift calculations |
| Test/control evidence available? |  |  | Required for upside category lift |

## Retail media guardrail

Monetized retail media belongs in Layer A when there is an active or planned monetization program. Retail-media addressable audience value belongs in Layer B when valuing the strategic audience asset. Do not count both for the same audience unless clearly separated.

---

# 7. Loss prevention, margin protection, and returns inputs

These inputs support the A6 Margin Protection and Connected Loss Prevention module.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Annual revenue in LP scope |  |  | Usually same as revenue in scope unless narrowed |
| Current shrink rate |  |  | Customer-provided preferred; otherwise use segment default |
| Annual shrink dollars |  |  | If known |
| Addressable shrink percentage |  |  | Use conservative/base; 73% should be ceiling only |
| Expected reduction of addressable shrink |  |  | Conservative/base/upside, capped in early-stage |
| Annual return value |  |  | Gross retail value of returned merchandise |
| Return fraud / abuse rate |  |  | Customer-provided preferred; otherwise benchmark range |
| Addressable return fraud percentage |  |  |  |
| Expected return fraud reduction |  |  | Capped in early-stage |
| SCO sales mix |  |  | % of revenue or transactions through self-checkout |
| SCO modality |  |  | Fixed SCO, Scan & Go, BYOD, mixed |
| Current SCO loss baseline |  |  | Required for upside assumptions |
| Annual LP investigations |  |  |  |
| Baseline investigation hours per case |  |  | Customer-provided preferred |
| LP labor rate |  |  | Fully loaded hourly rate |
| Investigation time-saved assumption |  |  | Capped in early-stage |
| Known phantom inventory / OSA issue? |  |  | Yes / No / Unknown |
| OSA-affected category sales |  |  | Required to enable OSA value |
| OSA baseline |  |  | Required for base/upside OSA value |

## LP guardrails

- Total A6 modeled Layer A value should not exceed 60-70% of the customer-confirmed addressable loss pool unless the user explicitly provides distinct, non-overlapping categories.
- SCO and phantom inventory / OSA submodules default OFF unless the customer provides required inputs.
- Detected fraud does not equal recovered cash.
- Investigation labor savings and shrink reduction can overlap.
- Safety belongs in Layer C unless customer provides cost data.

---

# 8. Workforce, labor, and store operations inputs

These inputs support intelligent labor flow, workflow protection, ESLs, task automation, and device reliability.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Fully loaded associate hourly wage |  |  |  |
| Average associates per store |  |  | Optional |
| Routine store inquiries per store per week |  |  | Associate interruptions |
| Minutes per routine inquiry |  |  |  |
| Queue abandonment / walkaway rate |  |  | Optional; use conservative if unknown |
| At-risk queue revenue |  |  | Use only if customer has data or scoped estimate |
| Weekly manual price-change hours per store |  |  | For ESLs / task automation |
| Annual pricing error cost per store |  |  | Customer-provided preferred |
| ESL deployment in scope |  |  | Yes / No / Planned |
| Device-related incidents per year |  |  | POS, handhelds, scanners, cameras, ESLs, signage, etc. |
| Average lost time per device incident |  |  | Minutes or hours |
| Cost per support ticket |  |  |  |
| Cost per truck roll |  |  |  |
| Annual truck rolls |  |  |  |

## Labor guardrail

Labor time saved is not automatically payroll savings. The TEV model can count workflow value, but if the user asks for hard savings, a separate realization factor is required.

---

# 9. Facilities, smart lighting, and energy inputs

These inputs support connected facilities and smart lighting, including PoE-powered lighting where relevant.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Annual utility spend |  |  |  |
| Lighting share of utility spend |  |  | % of utilities attributable to lighting |
| Stores in smart lighting scope |  |  |  |
| Current lighting type |  |  | LED, fluorescent, mixed, unknown |
| Controls / occupancy sensing in place? |  |  | Yes / No / Planned |
| Expected lighting energy savings |  |  | Apply only to lighting energy base |
| Facilities labor hours saved |  |  | Optional |
| Facilities labor rate |  |  | Optional |

## Energy guardrail

Apply smart-lighting savings to lighting-related energy, not total utility spend, unless the customer provides a broader building-controls baseline.

---

# 10. IT resilience, observability, and downtime inputs

These inputs support secure connected store foundation, observability, device reliability, downtime protection, and protected value at risk.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Annual store-impacting IT incidents |  |  | Network, application, API, cloud, device, payment, POS, etc. |
| Current MTTR |  |  | Mean time to resolution |
| IT labor cost per hour |  |  | Fully loaded |
| Annual network incidents |  |  |  |
| Annual outage hours |  |  | If known |
| Downtime exposure per hour |  |  | Customer-provided preferred |
| Revenue / productivity exposure per hour |  |  | Optional |
| Annual truck rolls / dispatches |  |  |  |
| Cost per truck roll |  |  |  |
| Observability maturity |  |  | Low / medium / high |
| Current tools |  |  | Splunk, ThousandEyes, AppDynamics, Meraki, etc. |
| Critical workflows dependent on network |  |  | POS, BOPIS, ESLs, signage, retail media, apps, contact center, etc. |

## Downtime guardrail

Downtime benchmark figures should be used as Layer C context unless the customer provides specific downtime exposure. Do not use broad cross-industry downtime benchmarks as hard defaults.

---

# 11. Cybersecurity, zero trust, and segmentation inputs

These inputs support cyber protected value, segmentation, identity security, and risk reduction.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Estimated cyber exposure |  |  | Customer risk model preferred |
| Annual incident probability |  |  | Customer-provided or benchmark-informed |
| Risk reduction assumption |  |  | Conservative/base/upside |
| Current segmentation maturity |  |  | Low / medium / high |
| Guest / associate / IoT / payment networks segmented? |  |  | Yes / No / Partial |
| MFA / identity protection maturity |  |  | Low / medium / high |
| Third-party / vendor access exposure |  |  |  |
| Ransomware exposure |  |  | Optional |
| Account takeover / credential-stuffing exposure |  |  | Optional |

## Cyber guardrail

Use customer-specific cyber exposure when available. If using benchmark exposure, anchor against published breach-cost ranges and disclose the source. Do not produce exposure values materially above published retail or cross-industry benchmarks without customer-provided justification.

---

# 12. Workplace safety inputs for Layer C

These inputs support optional C4 Workplace Safety Protected Value.

| Input | Customer value | Status | Notes |
|---|---:|---|---|
| Annual safety / violence incidents |  |  | Customer-provided required for numeric output |
| Workers' compensation cost |  |  | Optional |
| Legal / claims cost |  |  | Optional |
| Insurance premium exposure |  |  | Optional |
| Turnover linked to safety concerns |  |  | Optional |
| Cost per associate replacement |  |  | Optional |
| Safety technology in scope |  |  | Cameras, alerts, panic buttons, de-escalation, monitoring, etc. |

## Safety guardrail

Do not produce a numeric C4 Workplace Safety value unless the customer provides at least one cost input: workers' comp, claims, turnover, insurance, or legal exposure. Otherwise, present safety as qualitative protected value.

---

# 13. Scenario settings

Use these settings to control conservatism and output interpretation.

| Setting | Selected value | Notes |
|---|---|---|
| Scenario case | Conservative / Base / Upside | Default to base, but show all three when useful |
| Deployment scope | Pilot / Regional / Enterprise / Custom |  |
| Source strictness | Internal / Executive / customer-facing planning / Formal review | Formal review requires verified source sweep |
| Layer B visibility | Full / Summary / Appendix | For skeptical finance audiences, show summary with strong disclaimer |
| Layer C visibility | Full / Summary / Appendix | Protected value should remain separate from Layer A |
| Finance sensitivity requested? | No by default | Only if user explicitly asks |

---

# 14. Minimum input set for a quick first-pass TEV

If the customer only has limited data, collect these first:

| Priority | Input | Customer value | Status |
|---:|---|---:|---|
| 1 | Annual revenue |  |  |
| 2 | Store count |  |  |
| 3 | Average basket or annual transactions |  |  |
| 4 | Annual store visits |  |  |
| 5 | Known / identified customer rate |  |  |
| 6 | Annual customer contacts |  |  |
| 7 | Shrink rate or shrink dollars |  |  |
| 8 | Annual return value or return rate |  |  |
| 9 | SCO sales mix |  |  |
| 10 | Fully loaded associate labor rate |  |  |
| 11 | Annual utility spend |  |  |
| 12 | Annual IT incidents / downtime |  |  |
| 13 | Existing retail media program? |  |  |
| 14 | Deployment percentage / stores in scope |  |  |

---

# 15. Output preferences

| Output element | Include? | Notes |
|---|---|---|
| Executive TEV summary | Yes |  |
| Layer A direct annual impact | Yes |  |
| Layer B strategic asset proxy | Yes / Summary / Appendix | Always range + disclaimer |
| Layer C protected value | Yes / Summary / Appendix | Keep non-additive |
| Layer D future option value | Qualitative by default |  |
| Top 5 value drivers | Yes |  |
| Source IDs and caveats | Yes |  |
| Assumptions table | Yes |  |
| Follow-up discovery questions | Yes |  |
| Finance sensitivity | No by default | Only if explicitly requested |

---

# 16. Questions for the next discovery conversation

Use these after the first-pass TEV output to improve confidence.

## CX / identity

1. What percentage of in-store visits can currently be tied to loyalty, app, Wi-Fi, payment token, email, phone, or CRM identity?
2. Are store behavior signals currently used for personalization, retargeting, or retail media?
3. Are SMS, RCS, WhatsApp, or other messaging channels consented and active?

## Store operations

1. Which store workflows create the most associate friction today?
2. Where are queue, pickup, returns, or service-flow issues most visible?
3. How much time is spent on manual pricing, signage, or shelf-label updates?

## Loss prevention

1. What is the current shrink rate by format or category?
2. What portion of shrink is believed to be external theft, internal theft, process error, return fraud, or self-checkout loss?
3. What current LP systems are in place: cameras, POS exception reporting, case management, AI video, EAS, RFID?
4. How many investigations are performed per year, and how long do they take?

## IT / resilience

1. Which store workflows are most impacted when the network, app, or cloud service fails?
2. What are the most common incident types?
3. What is current MTTR?
4. What telemetry gaps make troubleshooting slow today?

## Cyber / trust

1. How are guest, associate, IoT, payment, and vendor networks segmented today?
2. What identity controls are in place for associates and third parties?
3. Has the customer quantified cyber exposure or downtime risk?

---

# 17. Final notes for the AI running the TEV calculator

When using this intake template with the Cisco Retail 2.0 TEV Calculator:

1. Do not over-ask before producing a first-pass scenario.
2. Use conservative defaults when inputs are unknown.
3. Clearly label which inputs are customer-provided versus benchmark defaults.
4. Keep Layer A, Layer B, Layer C, and Layer D separate.
5. Do not call TEV output ROI, profit, payback, NPV, or TCO.
6. Always disclose Layer B as strategic asset proxy value, not revenue.
7. Do not add Layer C protected value into Layer A direct impact.
8. Do not activate optional submodules unless required inputs are available.
9. Cite source IDs and caveats in outputs.
10. Recommend a Cisco Value Analyst engagement when outputs are large, high-stakes, or intended for formal business-case use.