How we score
Every score on this site, the free scan, the per-store reports and the leaderboard, comes from the same crawl and the same rubric. This page is that rubric, in plain words. No login, no admin access, nothing private.
A snapshot, not a promise
01 / Public only
We fetch only what is publicly there: the same pages, the same declared bot user-agents, an AI crawler would see. Raw HTML, no JavaScript.
02 / One rubric
Every score comes from the same deterministic 0 to 100 rubric. Same inputs, same score. Nothing is randomized and nothing is weighted by who is paying.
03 / A snapshot
A score reflects what we found the moment we crawled. Stores change constantly, so re-run a scan whenever you want a current read.
The same signals AI assistants use to decide whether to recommend you
The score is built from six categories, checked against exactly what an AI crawler can see. Structured data is the heaviest of them: it is the format AI assistants and search engines read first.
AI crawler access
GPTBot, ClaudeBot, PerplexityBot, Google-Extended and Bingbot: blocked or allowed, per bot, from your robots rules and the responses they actually get.
Structured data
Product JSON-LD completeness: name, price, availability, brand, images, identifiers. Missing or partial schema is the most common gap we find.
Titles and meta
Meta titles and descriptions that stand on their own, per page, instead of one template repeated across the catalog.
Social and canonical tags
Open Graph tags and canonical URLs that keep citations and shared links pointing home.
Image alt text
Coverage of descriptive alt text, which is what AI uses to understand product imagery.
Content depth
Description depth on product pages. Thin pages give an AI assistant nothing to quote, so they get skipped in answers.
Reported alongside the score
llms.txt
The emerging standard telling AI agents what your store sells, found or missing.
Catalog discovery
Whether an agent can find your products and collections from a cold start: sitemaps, their freshness, and whether robots.txt declares them.
JavaScript dependence
Content that only exists after JavaScript runs is invisible to AI crawlers. We read the raw HTML, the way they do.
One rubric, three surfaces
The free scan
Enter a store address, get a score in about fifteen seconds, then a deeper page-by-page report of a sample of product and collection pages. No account, no email required. Run it.
Per-store reports
Every store we have scanned has a public report at /scan/your-domain, with its score by category, its weakest pages, and the change since the last scan. A Markdown mirror sits at the same address with .md on the end, for agents.
The leaderboard
A curated list of well-known online stores, scored with the same crawl and ranked side by side. If a store is not listed, it is not because it failed; the list is curated. See it.
Scores are automated estimates from a store's public pages at one point in time, not a rating, a certification, or the last word on any brand. Think we got yours wrong, or want it removed? Contact us or read the full terms.
Can AI actually see your store?
Find out in about fifteen seconds. Free, no signup, any platform. Then fix what it finds: every missing point above has a fix in CortexCommerce.