Generic AI strategies fail. Every industry has unique:
The enterprise architects who succeed don't apply one-size-fits-all AI playbooks. They develop vertical-specific strategies that account for industry realities.
In this module, we'll use retail and commerce as our vertical—but the frameworks apply to any industry.
Three major protocols are shaping agentic commerce:
| Protocol | Provider | Purpose |
|---|---|---|
| ACP (Agentic Commerce Protocol) | OpenAI | Enables ChatGPT to browse and purchase on behalf of customers |
| AP2 (Agent Payments Protocol) | Lets Gemini make purchases with user-defined "mandates" | |
| MCP (Model Context Protocol) | Anthropic | The "USB port" connecting LLMs to enterprise systems |
Strategic implication: Retailers who implement these protocols become discoverable by AI shopping assistants. Those who don't become invisible.
Understanding key systems is essential for integration planning:
| System | Full Name | Role in AI Strategy |
|---|---|---|
| PIM | Product Information Management | Source of truth for product data AI agents reference |
| CDP | Customer Data Platform | Provides customer context for personalization |
| OMS | Order Management System | Enables agents to check orders, process returns |
| ATP | Available to Promise | Real-time inventory data for accurate availability |
| CMS | Content Management System | Houses editorial content AI agents cite |
Critical insight: AI agents are only as good as the data they access. Poor PIM data = poor recommendations. Outdated inventory = broken promises.
A complete vertical playbook addresses:
1. Opportunity Assessment
2. Data Readiness
3. System Architecture
4. Regulatory Considerations
5. Competitive Dynamics
Product information quality matters enormously:
Real-time data is critical:
Customer context enables personalization:
Guardrails prevent brand damage:
| Dimension | Low Readiness | Medium Readiness | High Readiness |
|---|---|---|---|
| Product Data | Incomplete, inconsistent | Mostly complete, some gaps | Comprehensive, structured |
| Inventory Data | Batch updates | Near-real-time | True real-time ATP |
| Customer Data | Siloed systems | Partially unified | Full CDP integration |
| API Infrastructure | Limited, legacy | Modern APIs, some gaps | Comprehensive, documented |
| Content Authority | Thin, generic | Some expert content | Authoritative, cited |
Before building a vertical playbook, understand the technology layers:
Each layer has specific design considerations for your vertical.
AI opportunities in retail cluster into five domains:
1. Discovery & Recommendation
2. Customer Service
3. Operations & Supply Chain
4. Merchandising
5. Content & Marketing
VERTICAL AI PLAYBOOK: [Industry/Segment]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
VERTICAL CONTEXT
Industry dynamics: [Key characteristics]
Customer expectations: [What customers expect from AI]
Regulatory environment: [Key constraints]
Competitive landscape: [Who's leading, what's table stakes]
OPPORTUNITY PRIORITIZATION
┌──────────────────────┬──────────┬───────────┬──────────┐
│ Opportunity │ Value │ Readiness │ Priority │
├──────────────────────┼──────────┼───────────┼──────────┤
│ │ H/M/L │ H/M/L │ 1-5 │
│ │ │ │ │
│ │ │ │ │
└──────────────────────┴──────────┴───────────┴──────────┘
DATA READINESS
| Data Domain | Quality | Gap | Remediation |
|-------------|:-------:|-----|-------------|
| Product | 1-5 | | |
| Customer | 1-5 | | |
| Inventory | 1-5 | | |
| Content | 1-5 | | |
SYSTEM ARCHITECTURE
[Diagram or description of integration approach]
REGULATORY REQUIREMENTS
• [Requirement 1]
• [Requirement 2]
• [Requirement 3]
COMPETITIVE STRATEGY
Differentiation focus: [Where to win]
Table stakes: [Must-have capabilities]
Watch list: [Emerging threats]
ROADMAP OVERVIEW
Phase 1 (Q1): [Focus]
Phase 2 (Q2): [Focus]
Phase 3 (Q3-Q4): [Focus]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Scenario: You're the enterprise architect for a specialty outdoor gear retailer:
Part 1: Opportunity Assessment
Rank these opportunities for your vertical (1=highest priority):
| Opportunity | Priority (1-5) | Why? |
|---|---|---|
| Conversational product advisor | ||
| Customer service automation | ||
| Demand forecasting | ||
| Product description generation | ||
| Review analysis for merchandising |
Part 2: Data Readiness Assessment
For your top 2 opportunities, assess:
OPPORTUNITY 1: _________________________
Required Data:
• _________________________
• _________________________
• _________________________
Current Quality (1-5): ___
Gap Description:
_________________________________________
Integration Required:
• _________________________
• _________________________
Part 3: Architecture Sketch
For your top opportunity, outline:
AGENT ARCHITECTURE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Customer Touchpoint:
[Where will customers interact?]
Agent Capabilities:
• [Capability 1]
• [Capability 2]
• [Capability 3]
Required System Integrations:
• [System 1] → [What data/actions?]
• [System 2] → [What data/actions?]
Guardrails Needed:
• [Guardrail 1]
• [Guardrail 2]
Success Metrics:
• [Metric 1]
• [Metric 2]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Part 4: Competitive Analysis
Research (or hypothesize) how competitors are using AI:
| Competitor | AI Capability | Threat Level |
|---|---|---|
| REI | H/M/L | |
| Patagonia | H/M/L | |
| Amazon | H/M/L | |
| Direct-to-consumer brands | H/M/L |
In Module 6.3: Portfolio Strategy, you'll learn how to evaluate and prioritize across multiple AI initiatives, balancing innovation investments with operational improvements.
You've reached the end of this module. Review the checklist below to ensure you've understood the key concepts.