AI findability is how easily AI assistants can find your brand, name it in an answer and cite your pages, and for most brands it is low. In the AI Software Index, 9,978 buyer questions asked on ChatGPT and Gemini in August 2026 across 126 software categories, the typical brand was named in just 17.5% of the answers in its category, and a quarter of all brands never reached 10% in any category they compete in. Those are the brands AI named at all.
How findable are brands in AI answers?
We took every brand the engines named in the index, 1,825 brands across 2,646 brand-and-category pairs, and measured how often each was named in its category's answers.
| Measure | Result |
|---|---|
| Median share of answers naming a brand in its category | 17.5% |
| Brand-category pairs named in under 10% of answers | 27.1% |
| Brands that never reach 10% in any of their categories | 25.3% |
| Brand-category pairs named in half the answers or more | 22.2% |
| Brand-category pairs never named first in any answer | 46.1% |
AI Software Index, August 2026: 9,978 answers from ChatGPT and Gemini across 126 categories. The index records a brand only once it is named in at least 5% of a category's answers, so brands below that line, the least findable of all, are not in these figures.
Findability also depends heavily on the category. In customer data platforms, the median named brand appeared in 46.2% of answers; in CMS software, 37.5%; in API management, 36.9%. At the other end sit CRM software at 8.9%, and employee onboarding, bookkeeping and subscription billing software at 10.0% each. The same product quality can mean very different findability depending on how many rivals share the answers.
What is AI findability?
AI findability is the chain of three things that has to hold for an AI assistant to put your brand in front of a buyer, each of which can be measured on its own.
Reachable. AI crawlers can fetch your pages. ChatGPT-User fetches pages while writing an answer, OAI-SearchBot and PerplexityBot index them, and a robots.txt rule, a login wall or a page that only renders in JavaScript can shut them out. Server logs show which bots arrive and what they request.
Named. Answers to your buyers' questions mention you. This is the part most tools measure, usually as visibility: the share of answers that name you across the questions and engines you track.
Cited. Answers link to your own pages, so a buyer who wants to check the claim lands on your site rather than a competitor's review of you. The gap between the two is covered in mentioned vs cited by AI.
A brand can pass one test and fail the next. Many sites are fully reachable and rarely named; some brands are named often but almost never linked. That is why a single number can hide where the problem is. For a short definition, see AI findability in the glossary.
How is AI findability different from AI visibility?
AI visibility is one part of findability, the named part: how often answers mention you. Findability is the whole chain, including whether crawlers can reach your pages in the first place and whether answers cite them. A visibility score tells you the outcome; findability tells you which link in the chain to fix. If visibility is low and crawlers never visit, the fix is technical. If crawlers visit constantly and you are still rarely named, the fix is on the third-party pages answers are built from, such as comparison articles, review sites and Reddit threads. For which tools report a visibility score, see AI visibility score, explained.
How can you check your AI findability?
Check each part separately, in order.
- Reachable. Look at your server or CDN logs for AI crawlers (ChatGPT-User, OAI-SearchBot, PerplexityBot, ClaudeBot) and check robots.txt does not block them. If your key pages never appear in the logs, start there.
- Named and cited. Run the free AI citation checker. It asks 20 buyer questions on ChatGPT, Gemini and Google AI Overviews, 60 checks in all, with no signup, and shows how often you are named, how often your pages are cited, and which competitors appear instead.
- Over time. One check is a sample, because answers change from run to run. To see whether findability is improving, track the same questions daily; the AI visibility tracker does that across ChatGPT, Perplexity, Gemini and Google AI Overviews.
Compare your result with the AI Software Index figures: the median named brand appears in 17.5% of its category's answers. If you are named in fewer than one in ten answers for your core buyer questions, you are in the least findable quarter of the brands AI names at all.
What this data cannot tell you
The index measures two engines, ChatGPT and Gemini, in one month, so Perplexity, Claude and Google AI Overviews may rank brands differently. It only records brands named in at least 5% of a category's answers, so the true median across every company in a category is lower than 17.5%. And it measures being named, not being cited: the cited part of findability needs your own tracking.