How the WooCommerce AI-shopping census works
If you found LeyMishCensus/1.0 (+https://www.leymish.com/woocommerce/census.html) in your server logs, this page
explains it. LeyMish Labs makes a free WooCommerce plugin that checks whether AI shopping agents can read a store.
To find out how common the problems it checks for are, we looked at public signals on a sample of WooCommerce
stores, and we publish only totals: no store is ever named.
What it requests (at most 5 requests per store)
/robots.txt: to read the rules for nine AI crawler names (GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-SearchBot, Claude-User, PerplexityBot, Perplexity-User, Google-Extended).- The home page: to see whether the store is in "coming soon" or maintenance mode.
/wp-json/wc/store/v1/products?per_page=1: WooCommerce's public Store API, to check it answers with valid JSON (and without a byte-order mark), and to get the address of one product.- That one product page: to read its Product structured data (JSON-LD) for identifiers (GTIN, MPN, SKU) and brand.
/llms.txt: to see whether one exists.
How it behaves
- It says who it is. The census uses its own user agent above, never a browser's or an AI crawler's. The AI crawler results come from reading your robots.txt rules, not from pretending to be those crawlers.
- It obeys robots.txt. If your robots.txt disallows
LeyMishCensus(or all robots) for a path, the census skips that request. - It's slow on purpose: one request per second overall, at most five per store, and one pass.
- It reads only public pages. No logins, no forms, no carts, no checkout.
Where the stores came from
The sample is taken from the public WooCommerce showcase and its filter pages, which list stores that chose to be featured. That's a sample of well-kept stores, not of WooCommerce as a whole, and we say so wherever we use the numbers. The list is kept in our company repository with the date it was collected.
What we publish
Totals and percentages only (for example "x of 100 stores block at least one AI crawler in robots.txt"). Per-store results stay private and are never published, sold or used for outreach.
Opting out
Add this to your robots.txt and the census won't request anything else from your store:
User-agent: LeyMishCensus
Disallow: /
Questions: open an issue on the plugin's GitHub repository.