Methodology · v2.0
How the score works
Every category, weight, and detector is public. Nothing about how a store earns its score is hidden — that's the entire premise of a certification mark you can trust.
A store is crawled — homepage, up to 10 product pages, up to 5 category pages, and one of each informational page type found (FAQ, contact, policies, testimonials). 29 rule-based detectors run against the crawled content and produce signals. 8 category scorers convert those signals into points, summing to a total out of 100. No detector calls an LLM or makes a judgement call — every point is traceable to a specific, checkable fact about the site.
Reviews & Reputation
18 pts
Measures the presence and visibility of customer reviews from third-party platforms like Trustpilot and on-site product reviews. Strong review signals prove that real people have bought from and trusted the store — the single most powerful indicator of legitimacy for first-time visitors.
What we check
- Trustpilot / third-party review integration
- On-site product reviews
AI Readiness
15 pts
Checks whether the store is actually accessible and machine-readable to AI crawlers and assistants — robots.txt rules for known AI bots, presence of llms.txt, structured data depth, sitemap quality, content extractability, and canonical/meta hygiene. This measures the preconditions for AI visibility, not whether AI engines currently mention the store.
What we check
- AI crawler access (robots.txt)
- llms.txt
- Structured data depth
- Sitemap quality
- Content extractability
- Canonical & meta hygiene
- security.txt
- humans.txt
AI crawlers we check robots.txt access for
User Experience
13 pts
Evaluates how easily customers can navigate, get help, and find important information. A store that is hard to use or lacks an FAQ and reachable contact page loses customer confidence immediately — poor UX is often the reason visitors bounce before buying.
What we check
- FAQ page
- Contact accessibility
- Informational page depth
Business Transparency
13 pts
Scores the presence of essential legal and contact pages — return policy, privacy policy, terms, shipping info, phone number, and email. Stores that hide this information raise immediate red flags; transparent stores give customers confidence they can get help or a refund if something goes wrong.
What we check
- Policy pages (return, shipping, privacy, terms)
- Visible contact details
Product Credibility
11 pts
Assesses whether product pages are well-presented with sufficient images, video, and structured data. Thin product pages with one image and no description suggest a dropshipping setup with no real product expertise, which directly reduces conversion and trust.
What we check
- Product image quality & quantity
- Structured data (Product/Review schema)
- Product video
Technical Trust
8 pts
Verifies that the site uses HTTPS and displays recognised payment and security badges. These are table-stakes requirements: a site without SSL or payment logos fails even a cursory trust check from both customers and search engines.
What we check
- HTTPS / SSL certificate
- Payment method logos
- Security & trust badges
Version history
- v2.0
- Added the AI Readiness category (8 new detectors) and re-weighted all 8 categories to sum to 100.
- v1.0
- Initial 7-category model: reviews & reputation, user experience, business transparency, SEO & authority, product credibility, social proof, technical trust.
Frequently asked questions
What is an AI Trust Score, and how is it calculated?
An AI Trust Score is a 0–100 rating of how trustworthy an e-commerce store looks to both human shoppers and AI crawlers. It comes from 29 rule-based detectors across 8 weighted categories — reviews, AI readiness, transparency, and more — that check specific, verifiable facts about the site. No language model makes the call; every point traces back to something you can check yourself.
Read the full answerWhat does 'AI readiness' mean for an online store, and why does it matter?
AI readiness measures whether a store is actually accessible and machine-readable to AI crawlers and assistants — robots.txt rules for bots like GPTBot and ClaudeBot, an llms.txt file, structured data depth, and clean canonical URLs. It's a precondition for AI visibility, not a guarantee an AI engine currently mentions the store — but a store that fails these checks can't be found by AI systems even if everything else about it is trustworthy.
Read the full answerSee where your store lands
Join the waitlist for a free scan against this exact methodology the moment scanning opens.
Join the waitlist
Social Proof
10 pts
Checks for testimonials, press mentions, and social-proof copy ("Trusted by 10,000+ customers"). These signals lower buying anxiety by showing that others have already made the decision to buy, making it much easier for a new visitor to follow suit.
What we check