We Checked 50 Startup Sites for AI Readiness | Opere18

We Checked 50 Startup Sites for AI Readiness

November 12, 2026·5 min read·

AI search readiness is mostly a few checkable signals on your own site. We ran those checks on 50 startup websites and found the same gap almost everywhere: the basics are in place, the answer-ready structure is not.

Want your own result? Run the free AI Search Visibility Scorecard: 12 checks, two minutes, a score out of 100.

What we checked and how

We took 50 company domains from startups and small businesses Opere18 has researched, 49 of which loaded. For each homepage we checked five signals: whether the site publishes any JSON-LD structured data, whether it declares an Organization or LocalBusiness entity, whether the homepage carries FAQPage markup, whether the site serves an llms.txt file, and whether robots.txt blocks any major AI crawler (GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, anthropic-ai, PerplexityBot, Google-Extended or CCBot).

This is a small sample of mostly AI and software startups, not a random sample of all businesses. Treat the percentages as a snapshot of this group, not a market-wide statistic. The checks are automated and read the homepage and robots.txt only.

The results

SignalSites passingShare of 49
Any JSON-LD structured data3367%
Organization or LocalBusiness schema3163%
llms.txt file served2449%
FAQPage markup on the homepage918%
Blocks at least one AI crawler in robots.txt00%

What stands out

Nobody blocked the AI crawlers. In this group, access is not the problem. The gap is structure: only 9 of 49 sites mark up questions and answers on the homepage, even though direct question-and-answer content is what answer engines quote.

About a third of sites still publish no structured data at all, which leaves an assistant to guess what the business is. And while llms.txt is far from universal, nearly half of these startups already serve one, so it has moved from curiosity to common practice in this segment.

What to do with this

  1. 01Add Organization schema with your exact business name, URL and sameAs profile links.
  2. 02Turn your real customer questions into FAQ sections with the answer in the first sentence, then mark them up as FAQPage.
  3. 03Keep AI crawlers allowed in robots.txt unless you have a specific reason to block them.
  4. 04Publish a short llms.txt that states what you do, who for and where your key pages are.
  5. 05Check the same facts match on your site, LinkedIn and directory listings. Our AEO audit checklist walks through it.

If you want a second pair of eyes on your AI search setup, our AI Search service covers the full audit and the fixes.

FREQUENTLY ASKED

How many startup sites did you check?

We checked 50 company domains and 49 loaded. They are mostly AI and software startups that Opere18 has researched, so the results describe this group, not all businesses.

What share of the sites had FAQ schema?

Nine of 49, or 18 percent, had FAQPage markup on the homepage. This is the biggest gap, since answer engines favour direct question-and-answer content.

Did any site block AI crawlers?

None of the 49 blocked any of the eight AI crawlers we tested in robots.txt, so crawler access was not the problem in this sample.

What is llms.txt?

It is a plain text file at the root of a site that tells AI systems what the business does and where its key pages are. 24 of 49 sites in this sample served one.

RELATED SERVICES

Delivered in 22+ markets worldwide.

Want this done for your business?

Free audit. No pitch. 24-hour turnaround.

Get free audit →

PROOF THIS WORKS

MORE ARTICLES