AI Commerce University

AI Retrieval Signals: the 8 factors

AI shopping agents evaluate every store across 8 scored factors before making a recommendation. Understanding each factor and what to fix first is the technical foundation of AI Commerce Intelligence.

The framework

How the AI Commerce Score v3.0 is built

The AI Commerce Score is a composite 0 to 100 measurement across 8 factors. Each measures a specific dimension of how well AI shopping agents can understand, trust, and recommend a store. Version 3.0 added Semantic Visuals and Image Clarity as a new factor reflecting how AI vision systems now evaluate product images directly.

Industry benchmark: Across 6,789 scanned stores, the average AI Commerce Score is 45 out of 100. Zero stores score above 85. 52% score below 50. Most ecommerce infrastructure was built for humans and Google, not for AI recommendation systems.
The 8 factors

Each factor explained

01. Semantic Visuals & Image Clarity (15%)

AI vision systems now read product images, not just text. GPT-4o Vision and Apple Intelligence compare what is actually in a photo against surrounding text and alt attributes. A generic alt tag like "image_322.jpg" tells AI nothing. "Black leather Chelsea boot, size 8" is rich semantic signal. This factor is new in v3.0 and reflects how rapidly AI vision is becoming part of recommendation evaluation.

02. AI Structured Signals (15%)

Structured data is the primary language AI agents use to understand what a store sells, at what price, and how trustworthy it is. Covers JSON-LD Product, Offer, and AggregateRating schema, Organization schema for merchant identity, and llms.txt quality. 72% of stores are missing critical schema markup.

03. Core Technical & Interpretability (15%)

Can AI crawlers actually access and parse the store? Covers robots.txt configuration (are GPTBot and ClaudeBot blocked?), Core Web Vitals via PageSpeed API (LCP under 2.5s, CLS under 0.1), and DOM structure quality including heading hierarchy. A technically broken store is an invisible store.

04. AI Trust & Transaction Confidence (15%)

Machine-readable trust signals AI checks before routing a buyer to a store. Return policy in crawlable HTML (not a JS modal), checkout accessibility for autonomous agents, and legal identity transparency in footer or contact pages. 64% of stores fail this check.

05. Commerce & Feed Accuracy (15%)

Price visible in HTML without JavaScript, schema price and priceCurrency present, inventory availability consistent across feed and live page. When an AI agent tells a buyer a price and the checkout shows a different one, that agent has failed the buyer. 58% of stores fail commerce accuracy checks.

06. User Intent Match (10%)

AI matches stores to buyer intent semantically, not by keywords. Specific positioning wins. "Organic collagen supplements for joint recovery in endurance athletes" matches very differently from "wellness products." Measured via LLM semantic embedding comparison against actual buyer prompt language in that niche.

07. Recommendation Confidence (10%)

The social proof signals AI can actually parse. AggregateRating schema with a real ratingValue and reviewCount from a verified review platform. Sentiment analysis on accessible review content. AI systems use this to predict post-purchase satisfaction and adjust how confidently they recommend a store.

08. External Authority Signals (5%)

Brand mentions in external LLM indexes, Reddit discussions, Perplexity citations, and AI-readable databases. Small weight now, expected to grow significantly by 2027 as LLM training data expands. A store with zero external presence is treated as unverified regardless of on-site quality.

Where to start: AI Trust and Transaction Confidence and Commerce and Feed Accuracy fail in more than 60% of stores. Moving prices to server-rendered HTML and placing return policy in crawlable text are the two highest-impact fixes for most stores. Semantic Visuals is the newest factor and the most underoptimized.
Score zones

What your score means

85+
Highly Recommendable. AI confidently recommends you.
70-84
Moderately Recommendable. Stronger signals win more traffic.
Below 50
AI Visibility Risk. Excluded from recommendations entirely.
FAQ

Common questions

What are the 8 AI retrieval signals?
The 8 factors in AI Commerce Score v3.0 are: Semantic Visuals and Image Clarity (15%), AI Structured Signals (15%), Core Technical and Interpretability (15%), AI Trust and Transaction Confidence (15%), Commerce and Feed Accuracy (15%), User Intent Match (10%), Recommendation Confidence (10%), and External Authority Signals (5%). Maximum score is 100.
Which AI retrieval signal matters most?
All five 15% factors are equally weighted. AI Trust and Transaction Confidence and Commerce and Feed Accuracy fail most often across scanned stores. Fixing prices in HTML and making return policy server-rendered typically provide the fastest score improvement.
How is the AI Commerce Score calculated?
The AI Commerce Score v3.0 is 100 points across 8 factors. Above 85 is Highly Recommendable. Between 70 and 84 is Moderately Recommendable. Between 50 and 69 is Low Recommendation Confidence. Below 50 is AI Visibility Risk.

See your store's AI Commerce Score free

Apply what you just learned. Get your real AI Commerce Score across all 8 v3.0 factors in 30 seconds.

Free. No credit card. Results in 30 seconds.
Continue reading

Up next