Agentic commerce: is your catalog ready for AI shopping agents?
What is agentic commerce?
3 to 5 trillion dollars of retail spending will be redirected by AI agents by 2030, according to Gartner. This figure transforms a trend into a strategic imperative for every e-commerce business.
Agentic commerce refers to a model where autonomous programs — AI shopping agents — search, compare, negotiate, and purchase products on behalf of a consumer. The user defines a need (« a waterproof jacket, size M, delivered in 48h, max budget 150€ ») and the agent handles the entire journey.
What was once science fiction is now operational:
- ChatGPT Shopping is live on Etsy, Amazon, and hundreds of marketplaces — the user describes what they’re looking for, the agent displays product recommendations with prices and reviews
- Microsoft Copilot Checkout integrates Shopify, PayPal, and Stripe to finalize the purchase directly within the conversational interface
- Perplexity Shopping offers an augmented comparator that synthesizes product listings, reviews, and shipping conditions in a single response
According to a Shopify Enterprise study (2026), most e-commerce catalogs are still invisible to these agents. The reason: product listings are designed for human eyes, not for reasoning algorithms.
Why do 73% of shoppers already use AI in their buying journey?
Adoption is already massive. 73% of online shoppers use at least one generative AI tool during their buying journey (Salesforce, State of Commerce 2026).
Usage breaks down as follows:
- 45% to generate product ideas (« What gift for a beginner hiker? »)
- 37% to get customer review summaries synthesized in one sentence
- 32% to compare prices across multiple retailers
- 28% to check availability and delivery timeframes
Zero-click commerce: direct purchase within the LLM
The model is evolving toward zero-click commerce: the purchase is completed within the agent’s interface, without the consumer visiting the merchant’s site. Copilot Checkout illustrates this: product selection, size choice, payment, and confirmation — all in 3 conversational exchanges.
This paradigm shift has a direct consequence: 70% of consumers are comfortable with the idea of an agent finalizing a purchase on their behalf (McKinsey, Consumer Pulse Q1 2026). Trust in transactional AI is established.
For brands, the question is clear: will your catalog be the one the agent recommends, or the one it ignores?
How do AI agents read your catalog?
An AI shopping agent works differently from a traditional search engine. It reasons over data instead of simply indexing keywords. To reason, it needs structured, explicit, and complete data.
Schema.org: the common language of agents
Agents rely on Schema.org markup to understand your product listings. Three schema types are essential:
- Product: name, description, brand, SKU, image, category
- Offer: price, currency, availability, shipping conditions, stock status
- AggregateRating: average rating, review count, rating distribution
An agent matches this data against the user’s query. If your listing is missing an attribute (such as material or delivery timeframe), the agent favors a competitor that provides this information.
Structured listings vs free-form content
The most frequent pattern in e-commerce catalogs: product descriptions are rich in editorial content but poor in structured data. An AI agent prefers an attribute « material: GOTS certified organic cotton » over a 200-word paragraph that mentions cotton in passing.
The 5-axis Agent-Readiness Score
We developed a proprietary evaluation framework: the Agent-Readiness Score. It measures a catalog’s ability to be read, understood, and recommended by an AI agent. The score is based on 5 axes:
- Intelligibility (20 pts): clarity of HTML structure, Hn hierarchy, semantic markup
- Product attributes (25 pts): completeness of standardized attributes (size, color, material, weight)
- Product information (20 pts): unique description, FAQ, size guides, technical specifications
- Categorization (15 pts): logical tree structure, breadcrumbs, ProductGroup
- Transactional data (20 pts): price, stock, shipping, returns, payment methods in Offer schema
Most e-commerce sites leverage less than half the potential of their structured data, according to Shopify and Search Engine Land (2026). The opportunity is real for brands that structure their catalog first.
What are the 7 actions to become agent-ready?
Each action targets an axis of the Agent-Readiness Score. Applied together, they transform a passive catalog into one that AI agents actively recommend.
Every product listing must include Product markup with at minimum: name, description, image, brand, sku, gtin (EAN/UPC), category. AI agents use these fields to uniquely identify your product. A product with GTIN is identified 3.4x faster than a product with a name alone.
Size, color, material, weight, dimensions: these attributes must appear in the additionalProperty markup (PropertyValue). Use standardized values (ISO sizes, standard colors, SI units). An agent comparing products needs homogeneous data across sources.
Every product listing should include 3 to 5 Q&A pairs marked up in FAQPage. Agents extract these FAQs to answer specific user questions: « Is this jacket suitable for heavy rain? ». On our optimized sites, product FAQs generate +27% rich impressions.
Agents heavily weigh social proof. AggregateRating markup with ratingValue, reviewCount, and bestRating allows the agent to instantly compare your product against competitors. Individual reviews in Review schema enrich the summary the agent presents to the user.
The Offer markup is the backbone of transactional readiness. Include: price, priceCurrency, availability (InStock/OutOfStock), deliveryLeadTime, shippingDetails, hasMerchantReturnPolicy. An agent with all transactional data can recommend your product with confidence.
Multimodal agents (GPT-4o, Gemini) analyze images. Use high-definition images (min. 1200px wide) with alt text that describes the product precisely: « Men’s navy blue waterproof jacket – front view – adjustable hood ». Each variant (color, angle) deserves its own marked-up image.
The llms.txt file at the root of your site tells LLMs which pages are priorities and how to interpret your catalog. Complement it with a dedicated XML sitemap that lists your product pages with lastmod and priority. Agents follow these directives to prioritize their crawl.
Example llms.txt structure:
> Catalog of 2,400 outdoor products
## Product listings
– /products/ : complete index
– /sitemap-products.xml : products sitemap
## Main categories
– /categorie/manteaux/
– /categorie/chaussures/
– /categorie/accessoires/
What is the impact of the Google March 2026 Core Update on agentic commerce?
The Google March 2026 update reshuffled the deck. The signals it rewards precisely overlap with AI agent criteria: original data, proven expertise, fast user experience.
The winners: original data and expertise
Sites publishing original data (product tests, benchmarks, proprietary photos) gained an average of +22% visibility on commercial queries. AI agents also favor sources that demonstrate verifiable expertise — exactly what Google now rewards.
The losers: affiliate and generic content
Affiliate sites with low added value suffered a –71% impact on organic visibility. AI-generated content published as-is, found on many sites, is now systematically downranked. Only expert content, enriched with proprietary data, retains its positions.
Technical performance: critical LCP
The update reinforces the weight of Core Web Vitals. Sites with an LCP above 3 seconds lost an average of –23% traffic. For an AI agent, crawl speed is also a factor: a slow site is expensive to analyze, therefore deprioritized.
The convergence is striking: what Google values in 2026, AI agents already require. Optimizing for agentic commerce means optimizing for Google at the same time.
How to measure your agent-readiness score?
Mesurer avant d’optimiser : c’est la méthode qui génère les meilleurs résultats. The 5-axis Agent-Readiness Score fournit un diagnostic précis de votre catalogue.
The 5 analysis modules
Each module evaluates one axis of the score:
- Intelligibility module: analyzes HTML structure, heading hierarchy, ARIA landmarks, and DOM readability by an agent
- Attributes module: verifies PropertyValue completeness in Product markup — compares your coverage to industry standards
- Information module: evaluates description uniqueness, FAQ presence, size guides, and technical specifications
- Categorization module: audits the tree structure, breadcrumbs, ProductGroup consistency, and sitemap coverage
- Transaction module: verifies Offer completeness (price, stock, shipping, returns, payment)
The most underutilized areas
According to studies published by Shopify and Search Engine Land in 2026, the most common gaps in e-commerce catalogs are:
- Product attributes: rarely standardized in structured data (size, color, material remain in free text)
- Transactional data: price is present, but shipping and return conditions are missing from the Offer schema
- Product FAQ: very few listings include FAQPage schema, even though it is the format most cited by LLMs
- ProductGroup: nearly absent, even on catalogs with variants (size/color)
Consider a site with 2,000 products: if attributes are in free text and FAQs are absent, an AI agent can only objectively compare a fraction of the catalog. The rest will be invisible.
Tools to audit your markup
- Google Rich Results Test: validates your structured data and previews rich snippets
- Schema Markup Validator (schema.org): complete analysis of all Schema types present on a page
- Screaming Frog: large-scale crawl to verify Schema coverage across the entire catalog
- Google Search Console: Enhancements report to track markup errors over time
Measure your Agent-Readiness Score
Free audit of your catalog: Schema.org markup, product attributes, transactional data. Results in 48h.
Book a free auditFrequently asked questions about le commerce agentique
What exactly is an AI shopping agent?
An AI shopping agent is an autonomous program that acts on behalf of a consumer: it searches for products, compares prices, reads reviews, checks availability, and can finalize a transaction. It relies on structured data (Schema.org) and product feeds to make decisions.
Is my WooCommerce site compatible with agentic commerce?
WooCommerce generates basic Product markup, but detailed attributes (materials, standardized sizes, AggregateRating) are often missing. A comprehensive Schema.org audit identifies the gaps and achieves a high agent-readiness score.
Is a specific budget required to become agent-ready?
The 7 priority actions primarily rely on optimizing existing data: Schema.org enrichment, attribute normalization, llms.txt file creation. The investment is a structural effort rather than a media cost.
What is the difference between GEO and agentic commerce?
GEO (Generative Engine Optimization) optimizes your visibility in LLM responses. Agentic commerce goes further: the AI agent completes the entire purchase. GEO is a foundational building block of agentic commerce — mastering GEO is the first step toward an agent-ready catalog.
Do AI agents respect robots.txt?
Major agents (GPTBot, ClaudeBot, PerplexityBot) respect robots.txt. The llms.txt file complements this approach by indicating to LLMs which pages are priorities and how to interpret your catalog.
How long before seeing results after optimization?
Structured data is indexed within 2 to 4 weeks by Google. LLMs update their sources more quickly. The first qualified traffic gains typically appear within the first month after implementation.
Does agentic commerce only apply to B2C?
B2B is equally concerned. AI procurement agents already automate supplier comparison, stock verification, and price negotiation. B2B catalogs structured with Schema.org Product gain a measurable competitive advantage.
Where do e-commerce merchants stand on agent-readiness?
According to Shopify and Search Engine Land studies (2026), most e-commerce catalogs leverage less than half the potential of their structured data. Product attributes and transactional data are the two most underutilized areas. Brands that structure their catalog first gain a measurable lead.
