How readable are dealer websites to AI shopping assistants? Every site we monitor, ranked by website provider and by car brand on whether its homepage and inventory read without JavaScript, and on llms.txt, structured data and robots.txt. Updated nightly.
How the rankings work
Whether an AI assistant reading the web for a shopper can understand the dealership: its homepage, its inventory, and the files that tell AI what the site offers. Many AI fetchers read a page's text without running its JavaScript, so a site that only shows vehicles after scripts run can look empty to them.
The homepage reads without JavaScript; the inventory lists vehicles in the page text, or in an llms.txt inventory page; an llms.txt file is published; the homepage carries schema.org structured data; and robots.txt doesn't block AI crawlers. A site is AI-ready when every check that could run passes.
Each site gets an AI readiness score from 0 to 100 built from the same checks. A provider's place is the median score of its checked sites, so one unusual site can't move it. A provider needs at least 50 checked sites to be ranked.
Many dealer websites answer automated visitors with a bot-protection page. When that happens to our crawler, the site is left out of the figures rather than counted as failing, because a wall in front of our crawler doesn't tell us what an AI assistant would get. The site count beside each provider shows how many sites were checked.
DealerDataOne-Crawler identifies itself by name and reads each site the way a script-free AI reader would. Our crawler page explains what it checks and how to reach us.
Nightly, from each site's latest check. New dealers, provider changes and providers reaching 50 checked sites appear the next morning. No dealership is named; the rankings cover providers and brands only.
Common questions
Only if the vehicles are in the page text it receives. Many dealer sites load vehicles with JavaScript after the page opens, or answer automated visitors with a bot-protection page, and either way an assistant can come away with nothing to read. The inventory check in these rankings measures the first of those.
llms.txt is a plain-text file at the root of a website that tells AI assistants what the site offers and where to find it, much like a sitemap does for search engines. Few assistants look for it on their own yet, so treat it as groundwork rather than a fix; some dealer platforms already publish one for every site.
That's the dealer's call, but blocking them site-wide keeps assistants from reading the site when shoppers ask about it. Most dealer websites we check don't block any.
A provider needs at least 50 sites we could check. Some have fewer dealers, and some put most of their sites behind bot protection that turns our crawler away, which leaves too few checked sites to rank them fairly.
Make sure vehicles appear in the page text without JavaScript, publish an llms.txt, keep schema.org markup on the homepage, and don't block AI crawlers in robots.txt. Much of this sits with the website provider, so it's worth asking yours.
They overlap. Readable text, structured data and an open robots.txt help search engines too. The difference is that many AI fetchers don't run JavaScript the way Google's crawler can, so content that only appears after scripts run is the bigger gap for AI.
By submitting this form, you agree to receive communications from us. You can unsubscribe at any time.
