We Scanned 81 Small-Business Websites for AI Search Readiness. Here Is What We Found.

The short version

Between launch and August 2026, the UpRyzze AI Visibility scanner has graded 81 real small-business websites (excluding sites we built ourselves) on the same deterministic checks: can AI engines reach the site, can they understand what the business is, and is there anything worth citing when they get there.

The findings, per check, across those 81 sites:

• 89% never explicitly welcome AI crawlers. Only 11% of sites name crawlers like GPTBot or ClaudeBot in robots.txt. 79% have a permissive file that never mentions them, and 10% block them or have no robots.txt at all.
• 53% have no llms.txt, the plain-language briefing file AI systems can read to understand a business.
• 49% have no Organization schema. Half the sites never tell machines, in machine-readable form, who the business is.
• 64% have no FAQ schema, the question-and-answer format AI answers lift from most directly.
• 43% have heading-structure problems: no clear main heading, or several competing ones.
• 14% have no usable sitemap.
• The one bright spot: 99% of sites were technically readable. The content is there. The machines just are not being told what it means.

Why this matters now

When someone asks ChatGPT, Perplexity, Gemini, or Google’s AI Overviews “who should I hire,” the engine reads websites directly and names one to three businesses in its answer. Whether your business is one of them depends on exactly the signals measured above: crawl access, machine-readable identity, and citable content.

The data says most small businesses are failing that test not because their sites are broken, but because their sites are silent. A site that loads perfectly and says nothing a machine can quote is invisible in an AI answer, and the customer who asked never learns the business exists. We wrote about how those recommendations get assembled in How to Get Your Business Recommended by ChatGPT.

What the gaps actually are

The robots.txt gap (89%). A permissive robots.txt is not the same as an open door. Explicitly naming AI crawlers is the clearest signal a site can send that its content is available for AI engines to read and cite, and it is the check the fewest sites pass. It is also among the cheapest to fix: a few lines of text.

The identity gap (49% no Organization schema). Structured data is how a site says “this is a business, this is its name, this is what it does” in the language machines parse natively. Without it, engines guess from prose, and they do not recommend guesses. This gap is why two businesses with similar sites can get wildly different AI treatment.

The citability gap (64% no FAQ schema). AI answers are question-shaped, so question-shaped content gets lifted first. A site with real FAQ content marked up properly hands the engine ready-made answers with the business’s name attached. Two thirds of sites offer nothing.

The structure gap (43%). Heading structure is how machines parse what a page is about. Multiple competing main headings, or none, degrade everything downstream.

An honest caveat about the sample

Every site in this data belongs to a business that was curious enough about AI visibility to run a scan. That selects for owners who are already paying attention. The average small-business website, the one whose owner has never heard of an AI crawler, is almost certainly in worse shape than these numbers show. Treat these figures as a floor, not a ceiling.

Two checks in our scanner (about-page depth and blog freshness) ran on too few sites in this cohort to report honestly, so they are not included. We only publish buckets with at least 20 sites behind them.

What to do with this

If you run a small business, the checklist writes itself, in order of cost:

1. Name AI crawlers explicitly in robots.txt.
2. Add an llms.txt that describes your business in plain language.
3. Ship Organization schema so machines know who you are.
4. Turn your most-asked customer questions into real FAQ content with FAQ schema.
5. Fix your heading structure: one clear main heading per page.

Or measure first: the free AI Visibility Score runs the same checks on your site in about two minutes, and the $497 Audit adds live testing across five AI engines, including which competitor gets recommended instead of you. Every site UpRyzze builds ships with all five items above from day one; that is the point of the platform.

FAQ

How many small-business websites were scanned, and how?

81 real small-business websites, scanned by the UpRyzze AI Visibility scanner through August 2026. Each site ran the same deterministic checks covering crawlability, structured data, and content structure. Websites built by UpRyzze were excluded so the numbers describe the market, not our own work.

What percentage of small businesses block AI crawlers?

In our data, 10% of sites either block AI crawlers or have no robots.txt at all, and another 79% have a permissive robots.txt that never names an AI crawler. Only 11% explicitly welcome crawlers like GPTBot and ClaudeBot, which is the strongest signal a site can send that its content is available to AI engines.

How many small-business websites have structured data?

About half. 49% of the sites we scanned had no Organization schema at all, meaning they never tell search and AI engines who the business is in machine-readable form. 64% had no FAQ schema, which is the format AI answers lift from most directly.

Are these numbers representative of all small businesses?

They likely understate the problem. Every site in this data belongs to a business curious enough about AI visibility to run a scan, which selects for owners already paying attention. The average small-business website is probably in worse shape than these numbers show.