Can AI even read your restaurant?

It runs twelve technical checks against your website: whether AI crawlers are allowed in, whether your homepage carries restaurant structured data, whether a machine can read your menu, and the supporting basics such as sitemap, meta description, and canonical tags. It reads your site the way an assistant does and scores what it finds out of one hundred.

No, it is the starter. The checker tests whether AI can read your website. The full audit tests whether AI actually recommends you: we ask dozens of real dining questions across the main AI assistants and score you against the restaurants you compete with for the same table. The checker takes a minute, the audit takes a specialist.

Because AI assistants cannot reliably read PDF menus. A PDF is a picture of your menu as far as a machine is concerned, so your dishes, prices, and dietary options are invisible at the exact moment a diner asks about them. Publishing the menu as a normal web page fixes it, and you can keep the PDF for print.

It means AI can read you, which is the entry requirement, not the win. Assistants build recommendations from your site plus reviews, listings, and mentions across the web, weighed against every competitor nearby. A strong score here with weak signals elsewhere still loses tables. That second half is exactly what the full audit measures.

The checker fetches a handful of pages from your site once, at the moment you run it: the homepage, robots.txt, the sitemap, and your menu page if we find one. It does not crawl beyond that and nothing runs on a schedule. If you unlock the detailed results, we receive your email and the score so we can follow up, and that is all.