Checked 2,474 local business sites for AI search readiness, curious what you're seeing with clients

[marketingandai]

I ran a crawl of 2,474 independent small business websites around Houston to see what AI search tools can actually read on them, and a couple of things surprised me so I'm curious if others are seeing the same.

• Only 31% passed four basics (AI search crawlers allowed in robots.txt, at least 120 words readable without JavaScript, business schema, one H1)
• Almost nobody blocks AI on purpose. 1.6% in robots.txt, but when I tested sites that allow the bots, about 3.8% more blocked them at the firewall (mostly 403s from bot protection), so roughly 5% overall
• 45% have no business schema at all, so nothing on the site confirms the NAP the citations are pointing at
• 19% show almost no text before JavaScript runs. For restaurants it's a third of them, which to GPTBot/ClaudeBot/PerplexityBot is basically a blank page
• 30% already have an llms.txt, but a lot of them are clearly generated by an SEO plugin

Two questions for the group:

1. Are you seeing client sites get blocked at the CDN/firewall level without the owner knowing? I hadn't been checking for it until now.
2. For local, is anyone tracking whether the JS-heavy sites (builder sites that load everything with scripts) are getting left out of AI answers more than others?

Happy to share more detail on how I tested any of these if it's useful.

 

[JS Girard]

llms.txt means literally nothing, as proven by every single test made about it. And No AI ingests schemas as anything else than base text, so I don't why either of those would be relevant in any shape or form to "AI readiness".