Agentic Commerce: Prepare Your Catalog for AI Shopping Agents Era

Agentic commerce is no longer a futuristic vision. It's an operational reality in the United States since September 2025, with real transactions occurring daily via ChatGPT and Perplexity.

According to McKinsey, agentic commerce could represent up to $1 trillion in B2C revenue in the United States by 2030, and between $3-5 trillion globally. BCG and Adobe report +4,700% growth in retail traffic from AI agents in one year.

At Aggil, we already build multi-agent systems that automate complex workflows. What agentic commerce standardizes today, we've been developing in production for several years. This article explains how to transform your catalog to be discoverable by AI agents, and why now is the time to act.

Understanding Agentic Commerce: Beyond the Buzzword

What Agentic Commerce Is NOT

Let's clarify this essential point first. Agentic commerce is not:

• A new advertising channel you'll "activate" tomorrow morning
• An improved chatbot on your e-commerce site
• Technology that replaces your marketing or sales teams
• A passing trend of generative AI

It's an infrastructure that fundamentally transforms how consumers discover and purchase products online.

What Agentic Commerce Really Is

Agentic commerce enables autonomous AI agents to act on behalf of users to:

1. Discover products by understanding intent in natural language

2. Compare options according to complex and personalized criteria

3. Negotiate the best conditions (price, delivery, warranties)

4. Purchase directly without leaving the conversational interface

Concrete example: A user asks ChatGPT: "I need a gift for my 10-year-old daughter's birthday who loves science, $50 budget".

The AI agent:

• Analyzes intent (educational gift, age range, interest area)
• Browses thousands of catalogs in seconds
• Recommends 3-5 relevant options with justifications
• Enables one-click purchase without leaving ChatGPT

What users experience as "magic" relies on complex technical infrastructure: standardized communication protocols (ACP, MCP), secure API connections, catalog optimization, and advanced recommendation models.

Market Status: Where Is Agentic Commerce in November 2025?

Currently Operational Platforms

ChatGPT Instant Checkout ✅ Operational (US only)

Launch: September 29, 2025

Availability: United States, all tiers (Plus, Pro, Free)

Merchants: Etsy (100% US sellers), 1M+ Shopify merchants, Walmart

Protocol: Agentic Commerce Protocol (ACP) open-source, developed with Stripe

Volume: 700 million weekly users

How it works for users:

1. Ask a shopping question in natural language

2. ChatGPT presents 3-5 relevant products

3. Click "Buy" for compatible products

4. Complete checkout in 30-90 seconds within the conversation

For merchants:

• Integration via Shopify or Etsy (automatic)
• Others: apply via OpenAI + Stripe/ACP integration
• Transaction commission (confidential amount)

Perplexity Buy with Pro ✅ Operational (US only)

Launch: November 2024

Availability: United States, Pro subscribers only ($20/month)

Merchants: Shopify + free Merchant Program

Unique innovation: "Snap to Shop" (visual search by photo)

Model: Currently no commission, free shipping for Pro subscribers

Particularity: 100% free merchant program with:

• Increased visibility in recommendations
• Free API to integrate Perplexity on your own site
• Dashboard with shopping analytics and trends
• 5x increase in shopping queries since launch

Google AI Mode Shopping ⚠️ Announced, not yet deployed

Announcement: Google I/O, May 2025

Expected availability: "In the coming months" (US)

Features:

• Conversational shopping with "query fan-out" (simultaneous searches)
• Track Price + "Buy for Me" (automatic agentic checkout)
• Virtual Try-On ✅ (already available in beta)

Infrastructure: 50 billion listings, 2 billion updated every hour

How AI Shopping Agents Work: Technical Architecture

A Shopping Agent's Decision Pipeline

When a user asks ChatGPT or Perplexity for a product recommendation, here's what happens behind the scenes:

The Three Critical Selection Factors

For a product to be recommended by an AI agent, three conditions must be met:

1. Technical Accessibility (Infrastructure)

• Catalog connected via standard protocol (ACP, MCP, API)
• Machine-readable structured data (JSON-LD, Schema.org)
• Real-time availability and pricing
• Compatible checkout system (Stripe, Google Pay, etc.)

2. Semantic Relevance (Content)

• Rich and unique descriptions (not manufacturer copy-paste)
• Complete metadata (attributes, specifications, uses)
• Explicit usage context
• Natural language aligned with user queries

3. Quality Signals (Reputation)

• Consolidated and verified customer reviews
• Return rate and satisfaction
• Delivery speed and reliability
• Performance history (if data available)

What differs from traditional SEO: AI agents don't "crawl" your website. They directly query your catalog via API or read pre-indexed structured data. Visual optimization or internal linking have no impact. Only the semantic richness of your product data and their machine accessibility matter.

Preparing Your Catalog for AI Agents: Technical Guide

Phase 1: AI Discoverability Audit (2-3 weeks)

Technical Diagnosis

Infrastructure:

• ✓ Is your catalog accessible via REST/GraphQL API?
• ✓ Do you have Schema.org structured data on your product pages?
• ✓ Is your site pre-rendered (SSR) or 100% client-side (SPA)?
• ✓ Can AI agents read your content without JavaScript?

Critical test: Disable JavaScript in your browser. If your product pages disappear, they're invisible to AI agents.

Content:

• ✓ Are your product descriptions unique (≠ manufacturer text)?
• ✓ Do you have rich metadata (attributes, specifications, usage context)?
• ✓ Are essential information in structured text (not in images)?
• ✓ Do you have consolidated and actionable customer reviews?

Current discoverability test:

If you have US access or via VPN:

1. Query ChatGPT or Perplexity about your product category

2. Note if your products appear in recommendations

3. Analyze which competitors are cited and why

Audit deliverables:

• Technical compatibility matrix (infrastructure, data, formats)
• AI discoverability score by product category
• Gap analysis vs. referenced competitors
• Prioritized optimization roadmap

Phase 2: Catalog Optimization (4-8 weeks)

A. Technical Infrastructure

1. Pre-Rendering for AI Agents

If your site is SPA (React, Vue, Angular), implement a pre-rendering solution:

• Next.js SSR (Server-Side Rendering)
• Prerender.io or equivalent solutions
• Static generation for critical product pages

Validation: curl -A "GPTBot" https://yoursite.com/product-x must return complete HTML.

2. JSON-LD Structured Data

Implement Schema.org Product markup on each product page:

{
  "@context": "https://schema.org/",
  "@type": "Product",
  "name": "Product Name",
  "description": "Unique enriched description...",
  "brand": {"@type": "Brand", "name": "Brand"},
  "offers": {
    "@type": "Offer",
    "price": "99.99",
    "priceCurrency": "USD",
    "availability": "https://schema.org/InStock",
    "seller": {"@type": "Organization", "name": "Your Store"}
  },
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": "4.5",
    "reviewCount": "127"
  },
  "additionalProperty": [
    {"@type": "PropertyValue", "name": "Material", "value": "Organic cotton"},
    {"@type": "PropertyValue", "name": "Usage", "value": "Outdoor, all seasons"}
  ]
}

3. Real-Time Catalog API

Expose your product data via standardized REST API:

• Semantic search endpoint
• Filters by attributes, price, availability
• JSON responses with consistent structure
• Rate limiting and authentication

Recommended framework: Anthropic's Model Context Protocol (MCP) for maximum compatibility.

B. Semantic Enrichment

1. LLM-Optimized Product Descriptions

❌ Bad example (typical 2015 e-commerce):

Men's hiking shoes - Model XZ500
Color: Black/Gray
Available sizes: 8-12
Material: Synthetic

✅ Good example (AI agent optimized 2025):

Technical Men's Hiking Shoes XZ500 -
Designed for mountain trekking and difficult trails
 
Ideal for: 10-15 mile hikes, rugged terrain,
all seasons including light snow
 
Construction:
- Waterproof breathable Gore-Tex upper (100% waterproof guarantee)
- High-grip Vibram sole for wet rocks
- Ankle reinforcements for lateral support on slopes
- Protected toe for rocky descents
 
Long-distance comfort:
- EVA cushioning for impact absorption
- Mesh lining moisture evacuation
- Quick lacing cable tightening system
 
Suited for wide feet. True to size (no need to size up).
Break-in period: 2-3 outings.
 
Weight: 20.5oz per shoe (size 9)
Available sizes 7-13, colors Black/Gray and Brown/Beige

Key principles:

1. Explicit usage context ("for whom", "for what", "under what conditions")

2. Conversational natural language (as if advising a friend)

3. Differentiating attributes highlighted

4. Anticipation of frequent questions

5. Varied vocabulary to match different user formulations

2. Advanced Structured Metadata

Enrich each product with:

• Semantic categories: "Gifts for science-loving kids", "Winter trail running gear"
• Use cases: Explicit tags for situations (travel, daily, sports, professional...)
• User profiles: Who typically buys this product
• Complementary products: Intelligent semantic association
• Seasonality: Relevance according to time of year

C. Quality and Trust Signals

1. Customer Review Consolidation

AI agents analyze reviews to assess quality. Optimize:

• Volume: Minimum 15-20 reviews per product for credibility
• Freshness: Recent reviews (< 6 months) prioritized
• Structure: Rating + argued text > rating alone
• Processing: Extraction of key insights (recurring pros/cons)

2. Performance Data

If available, expose:

• Return rate (⬇️ = positive signal)
• Actual average delivery time
• Post-purchase satisfaction score
• Average product lifespan

Phase 3: Connection to Agentic Protocols (6-10 weeks)

ACP Integration (Agentic Commerce Protocol)

The ACP protocol, open-source since September 2025, allows AI agents to:

• Discover your catalog
• Check real-time availability and pricing
• Initiate and complete transactions

Integration steps:

1. Infrastructure Setup

# Clone official ACP repo
"text-blue-400">git clone https://github.com/openai/agentic-commerce-protocol
 
# Install dependencies
"text-blue-400">npm install @acp/merchant-sdk stripe

2. ACP Server Configuration

"text-purple-400">import { ACPMerchantServer } "text-purple-400">from '@acp/merchant-sdk';
"text-purple-400">import Stripe "text-purple-400">from 'stripe';
 
"text-purple-400">const acpServer = new ACPMerchantServer({
  merchantId: 'your-merchant-id',
  catalogApiUrl: 'https://api.yoursite.com/catalog',
  checkoutProvider: new Stripe(process.env.STRIPE_SECRET_KEY),
 
  // Webhook "text-purple-400">for agent notifications
  webhookSecret: process.env.ACP_WEBHOOK_SECRET,
 
  // Security configuration
  rateLimits: {
    catalogSearch: { requests: 100, window: '1m' },
    checkout: { requests: 10, window: '1m' }
  }
});
 
// Catalog search endpoint
acpServer.on('catalog.search', "text-purple-400">async (query) => {
  "text-purple-400">return "text-purple-400">await yourInternalAPI.searchProducts({
    query: query.text,
    filters: query.filters,
    maxResults: query.limit || 10
  });
});
 
// Availability check endpoint
acpServer.on('product.check_availability', "text-purple-400">async (productId) => {
  "text-purple-400">return "text-purple-400">await yourInternalAPI.getAvailability(productId);
});
 
// Order creation endpoint
acpServer.on('order.create', "text-purple-400">async (orderData) => {
  "text-purple-400">const order = "text-purple-400">await yourInternalAPI.createOrder(orderData);
 
  "text-purple-400">return {
    orderId: order.id,
    status: 'confirmed',
    trackingUrl: order.trackingLink
  };
});
 
acpServer.listen(3000);

3. Platform Registration

Once your ACP server is operational:

• Shopify: Activation via Shopify admin (Settings > Apps > ChatGPT Commerce)
• OpenAI Direct: Apply via https://openai.com/merchant-apply
• Perplexity: Merchant Program registration (free)

MCP Integration (Model Context Protocol)

For maximum compatibility with all AI agents (not just commerce):

"text-purple-400">from mcp "text-purple-400">import MCPServer, Tool
"text-purple-400">import json
 
server = MCPServer(name="your-catalog")
 
@server.tool()
async "text-purple-400">def search_products(
    query: str,
    max_results: int = 10,
    filters: dict = None
) -> list[dict]:
    """
    Semantic search "text-purple-400">in product catalog.
 
    Args:
        query: Natural language search
        max_results: Maximum number of results
        filters: Optional filters (price, category, etc.)
 
    Returns:
        List of products "text-purple-400">with enriched descriptions
    """
    results = await your_internal_api.semantic_search(
        query=query,
        filters=filters,
        limit=max_results
    )
 
    # Enrichment "text-purple-400">for agent context
    "text-purple-400">for product "text-purple-400">in results:
        product['_context'] = {
            'usage': product.get('usage_contexts', []),
            'for_who': product.get('target_profiles', []),
            'season': product.get('seasonality', 'all-year')
        }
 
    "text-purple-400">return results
 
"text-purple-400">if __name__ == '__main__':
    server.run(port=5000)

Phase 4: Continuous Optimization (Ongoing)

A. Discoverability Monitoring

Dashboard to implement:

Recommended tracking tools:

• Weekly manual tests via ChatGPT/Perplexity
• Automated mention scraping (tools like BrandWatch, Mention)
• Native platform analytics (when available)

B. A/B Testing on Descriptions

Methodology:

1. Test product selection: 20-50 products with significant search volume

2. Description variations: Version A (baseline), Version B (AI optimized), Version C (different focus)

3. Test period: 4-6 weeks minimum

4. Measurement: Track appearances and conversions by version

5. Iteration: Deploy winning version, new test

Variables to test:

• Description length (short 100 words vs long 300 words)
• Context positioning (beginning vs end)
• Tone (factual vs conversational)
• Structure (paragraphs vs lists)

The Aggil Approach: From Audit to Deployment

At Aggil, we support our clients in preparing for agentic commerce with a proven methodology from our AI projects since 2008.

Our 4-Phase Methodology

Phase 1: Strategic Audit (2-3 weeks)

Objective: Assess your maturity and define optimal roadmap.

Deliverables:

• Technical infrastructure audit (APIs, structured data, pre-rendering)
• Catalog quality analysis (description richness, metadata, reviews)
• AI discoverability competitive benchmark
• ROI estimate by product category
• Prioritized roadmap with identified quick wins

Our added value: We've already conducted this exercise for several sectors (fashion retail, electronics, home). We know which patterns work in production vs those that fail.

Indicative pricing: €8,000 - €15,000 depending on catalog size

Phase 2: Proof of Concept (4-6 weeks)

Objective: Validate feasibility and ROI on limited scope.

Scope:

• 1-2 priority product categories (50-100 references)
• Description optimization + structured data
• MCP/ACP server setup (test environment)
• Discoverability tests on US platforms (if access possible)

Deliverables:

• Optimized test catalogs
• Functional agentic protocol infrastructure
• Dashboard monitoring initial results
• Complete technical documentation

Our approach: Rapid PoCs with frequent iterations. A/B testing on descriptions to identify most performant patterns for your sector.

Indicative pricing: €15,000 - €30,000

Phase 3: Production Deployment (8-12 weeks)

Objective: Generalization to entire catalog with monitoring.

Deliverables:

• Complete catalog optimization (partial automation via AI)
• Production MCP/ACP servers with high availability
• Platform integrations (Shopify, OpenAI, Perplexity based on eligibility)
• Real-time monitoring and alerting system
• Internal team training (marketing, product, tech)
• Operational documentation and runbooks

Knowledge transfer: We don't deliver a "black box". Complete training of your teams for total autonomy on:

• Creating/optimizing new product pages
• Using monitoring tools
• Interpreting metrics and adjustments
• Maintaining protocol infrastructure

Indicative pricing: €35,000 - €80,000 depending on catalog size and complexity

Phase 4: Continuous Optimization (Ongoing)

Objective: Improve performance and adapt to evolutions.

Services:

• Monthly A/B tests on descriptions
• Platform watch and emerging best practices
• Scoring algorithm adjustments
• Technical support and strategic consulting
• Continuous training on new features

Our commitment: Active monitoring of protocol evolutions (ACP, MCP), participation in working groups, and sharing of learnings between our clients (anonymized).

Indicative pricing: €2,000 - €5,000/month depending on support level

What Differentiates Us

✅ Rare Multi-Domain Expertise

• 17 years in digital transformation and e-commerce
• Multi-agent systems in production since 2022
• Complex API integrations on 50+ projects
• Machine Learning applied to commerce (recommendations, pricing)

✅ Inside Protocol Knowledge

• Active contribution to MCP and ACP communities
• Real tests on US platforms (partnerships)
• Daily monitoring of standard evolutions
• Learnings from our first PoCs shared

✅ Pragmatic ROI-Driven Approach

• No "big bang": progressive deployment by waves
• Systematic before/after measurement on business KPIs
• Intelligent automation (AI to enrich, not replace humans)
• Guaranteed knowledge transfer (no dependency)

✅ Native Compliance & Security

• GDPR-compliant by design
• EU AI Act: governance and transparency
• CIR/CII accreditation: up to 30% tax reduction on our services

Agentic Commerce ROI: Projected Figures

Global Market Data

McKinsey (2025):

• $1 trillion USD US B2C revenue by 2030 via agentic commerce
• $3-5 trillion USD globally
• 44% of AI search users consider it as primary source

BCG / Adobe (2025):

• +4,700% retail traffic growth from GenAI YoY (July 2025)
• +32% time spent on site for visitors from AI agents
• +10% engagement rate for customers arriving via agents
• -27% bounce rate
• $26 billion USD AI search ad spending USA by 2029

Projected ROI for Average European E-commerce

Assumptions:

• Current revenue: €5M/year
• Organic Google traffic: 60% of revenue (€3M)
• European agentic commerce launch: Q2 2026
• AI agent market share: 10% year 1, 25% year 3

Conservative scenario (prepare now):

ROI: 17.6x over 3 years

Pessimistic scenario (wait-and-see, preparation in 2027):

Cost of waiting: -€985K over 3 years

Additional Operational Gains

Beyond revenue:

• -40% time spent on product page writing (AI-assisted)
• +25% catalog data quality (cleaning process)
• Reusability: structured data also useful for voice SEO, marketplaces, comparators
• Competitive advantage: internal expertise on emerging channel

Concrete Examples: Agentic Commerce in Action

Example 1: Personalized Surprise Gift

User: Marie, 34, looking for a birthday gift for her husband passionate about cycling.

Query to ChatGPT:

"My husband does road cycling, he already has a good bike and clothing. I'd like to offer him something useful for his long rides, budget $80-120. He often leaves early in the morning in autumn."

What happens:

1. Intent analysis:

- Profile: equipped road cyclist

- Need: useful accessory for long rides

- Context: cool autumn mornings

- Budget: $80-120

2. Catalog search:

Agent queries connected catalogs with semantic criteria:

- "Road bike accessories"

- "Long distance rides"

- "Cool autumn conditions"

- Excludes basic equipment (already owned)

3. Recommendations:

ChatGPT suggests:

Option 1: Waterproof frame bag with smartphone holder ($89)

"Perfect for long rides. Keeps energy bars and phone within reach, with touchscreen usable even with gloves. Waterproof for dewy mornings."

Option 2: Ultra-powerful rechargeable front light ($115)

"Ideal for pre-dawn departures in autumn. 8h battery in eco mode, 3h in full power. USB-C rechargeable."

Option 3: GPS computer with mapping ($105)

"To explore new routes without getting lost. Displays elevation, distance, turn alerts, and can follow predefined GPX tracks."

Each recommendation includes: contextualized justification, product page link, "Buy" button if Instant Checkout compatible.

Result: Marie buys the light in 2 clicks. Transaction in 45 seconds.

Why these products were chosen:

• Descriptions explicitly mentioned "morning rides", "autumn", "long distance"
• Metadata indicated "for equipped cyclists" (not beginners)
• Price within range
• Immediate availability
• Positive customer reviews (>4.5/5 with relevant mentions)

Example 2: Complex Multi-Criteria Search

User: Thomas, 28, moving into a Parisian studio.

Query to Perplexity:

"I'm looking for a desk for working from home, my studio is 180 sq ft. I need to store my laptop, 24-inch screen, printer, and files. Max budget $300. Modern/minimalist style."

What happens:

Constraint analysis:

• Limited space (180 sq ft)
• Multi-function: work + storage
• Equipment: laptop + 24" screen + printer + files
• Style: modern/minimalist
• Budget: ≤$300

Query fan-out (Perplexity):

Agent launches several parallel searches:

• "Compact desks small spaces"
• "Desks with integrated storage"
• "Dual monitor support desks"
• "Scandinavian minimalist desks"
• "Desks $200-300"

Perplexity recommends:

Desk 1: Compact desk with integrated wall shelves ($279)

"Dimensions 47x24in, ideal for 180 sq ft as uses vertical space. Integrated shelves for files and printer. Screen + laptop support. Clean light wood style."

Desk 2: Space-saving corner desk ($259)

"Optimizes room corner. 39x39in top, 24in side width. Drawer storage under top. Integrated cable management. 30min assembly."

Desk 3: Foldable wall-mounted design desk ($189)

"Folds away when not in use: saves 16in. 35x20in folding top. Mini-integrated shelves. Limit: no printer on top (weight), need separate furniture."

Each option with:

• 3D rendering or situational photo
• Dimension comparison vs 180 sq ft constraint
• Pros/cons list for specific use
• Relevant customer reviews (mentions "small space", "remote work")

User action:

1. Visually compares the 3 options

2. Asks: "Desk 1, how much weight can it support?"

3. Perplexity responds instantly: "Max top load: 110lb (largely sufficient for your setup). Shelves: 33lb each."

4. Thomas buys via "Buy with Pro" in one click

Total time query → purchase: 3 minutes

Why this desk was selected by the agent:

• Description explicitly mentioned "small space", "studio", "vertical optimization"
• Complete technical specifications (dimensions, supported weight, assembly)
• Metadata: "remote work use", "Scandinavian minimalist style"
• Situational photos "200 sq ft studio" in product page
• 47 customer reviews with recurring mentions "perfect small apartment"

Pitfalls to Avoid

1. Optimize to Manipulate (and Lose Trust)

Temptation: Use "strategic text sequences" (STS) to force agents to recommend your products at all costs, even if not optimal for the user.

Reality: Platforms detect and penalize these practices. OpenAI and Perplexity already have anti-gaming systems.

Our recommendation: Optimize for real relevance, not for gaming. AI agents become more sophisticated each month. What works today in manipulation will be banned tomorrow.

2. Neglect Data Accuracy

Common error: Incorrect prices, availability, or specifications in catalog.

Consequence:

• Recommendation followed by purchase impossibility = catastrophic experience
• User complaints escalate to platforms
• Downranking or merchant exclusion

Our approach: Mandatory real-time synchronization. Better not to be recommended than recommended with false data.

3. Underestimate Organizational Change

Reality: Preparing your catalog for AI agents isn't just a tech project. It impacts:

• Marketing: new way of thinking about product content
• Sales: less control over presentation (agent reformulates)
• IT: APIs and monitoring to maintain
• Support: new customer questions via these channels

Our method: Cross-functional training and internal ambassadors from Phase 2.

4. Want to Do Everything In-House Without Expertise

Classic pitfall: "We have a dev team, we'll do it ourselves."

Problem: Rare intersection of required skills:

• E-commerce + AI/ML
• NLP and semantics
• Emerging protocols (ACP, MCP)
• LLM optimization

Our value: We've already made the mistakes, identified working patterns, and can accelerate your timeline by 4-6 months.

The Opportunity Is Now: Act Before the Wave

Agentic commerce is not a futuristic vision. It's an operational reality in the United States since September 2025. First real transactions occur daily via ChatGPT and Perplexity. Google is coming in the next months.

In Europe, we have a unique 6-12 month window to:

1. Prepare our infrastructure while competition waits

2. Master optimization techniques before it becomes a commodity

3. Be among the first recommended during European launch

4. Build sustainable advantage on a channel that will represent 25-40% of online commerce by 2030

At Aggil, we don't sell dreams. We build AI systems in production for 17 years. Agentic commerce, we study it, test it, and implement it since the first protocols emerged.

Our Commitment

✓ Total transparency on what works (and what doesn't yet)

✓ No black box: complete knowledge transfer

✓ Measurable ROI with KPIs defined before project

✓ GDPR and EU AI Act compliance native

Next Steps: How to Get Started

1. Free AI Discoverability Audit

We offer a free 2-hour audit to:

• Assess current catalog maturity
• Test your discoverability on US platforms (if access)
• Identify 3 priority quick wins
• Estimate your potential ROI
• Define realistic roadmap

2. "Agentic Commerce 101" Workshop

4-hour training for your teams (marketing, product, tech):

• Understand how AI shopping agents work
• Decode ACP/MCP protocols
• Catalog optimization best practices
• Live demonstrations
• Q&A personalized to your sector

Price: €2,500 (up to 10 participants)

3. "Agent-Ready" PoC in 6 Weeks

Rapid validation on 1 product category:

• 50-100 optimized products
• MCP/ACP server setup (test)
• Discoverability tests
• Generalization recommendations

Price: €15,000 all inclusive

Conclusion: The Agentic Era Doesn't Wait

Agentic commerce redefines discovery and online purchasing as profoundly as Google redefined information search 25 years ago.

The numbers are clear:

• $1 trillion USD in US by 2030
• 700 million ChatGPT weekly users
• +4,700% retail traffic growth from AI agents in 12 months

The technology is ready:

• Open-source standardized protocols (ACP, MCP)
• Production platforms (ChatGPT, Perplexity)
• Secure payment infrastructure (Stripe, Google Pay)

The European market has a unique window:

• 6-12 months before European deployment
• Time to prepare infrastructure
• First-mover opportunity on a channel that will be critical

At Aggil, we've been building this capability since 2022 with our multi-agent systems. We have the technical expertise, proven methodology, and learnings from first deployments to support you.

The question isn't "if" agentic commerce will arrive in Europe.

The question is: will you be ready when it does?

Let's transform your catalog together into a strategic asset for the agentic era.

---

About Aggil

Since 2008, Aggil supports companies in their digital and AI transformation. Specialized in multi-agent systems, machine learning, and complex integrations, we help our clients automate their business processes with constant focus on measurable ROI and knowledge transfer.

Key expertise: Multi-agents | Catalog Optimization | Agentic Protocols (ACP/MCP) | E-commerce AI | EU AI Act Compliance

Accreditation: CIR/CII (up to 30% tax reduction on R&D services)

---

:::resources

Technical Documentation:

[Agentic Commerce Protocol (ACP) - GitHub](https://github.com/agentic-commerce-protocol/agentic-commerce-protocol)

[Model Context Protocol (MCP) - Anthropic](https://modelcontextprotocol.io)

[ChatGPT Instant Checkout - Merchant Guide](https://chatgpt.com/merchants/)

[Perplexity Shop like a Pro](https://www.perplexity.ai/hub/blog/shop-like-a-pro)

[OpenAI Commerce Documentation](https://developers.openai.com/commerce/)

Studies and News:

[Stripe & OpenAI - Agentic Commerce Protocol](https://stripe.com/blog/developing-an-open-standard-for-agentic-commerce)

[PayPal adopts Agentic Commerce Protocol](https://newsroom.paypal-corp.com/2025-10-28-OpenAI-and-PayPal-Team-Up-to-Power-Instant-Checkout-and-Agentic-Commerce-in-ChatGPT)

Aggil Services:

[Learn more about our services](/services)

[Contact us for a free AI audit](/contact)

:::