AI search visibility, measured

What AI search visibility means

Classic SEO asks whether you rank on page one. AI search asks a different question: when someone types "best CRM for a 10-person startup" into an assistant, is your product one of the four or five names in the answer?

There is no page two. A product that isn't named is invisible for that buyer, however good its Google rankings are.

How ShownInAI measures it

1. 01We infer your category and closest competitors from your website.
2. 02We write the prompts real buyers use: "best X for Y", "X alternatives", "X vs Y", "tools like X". None contain your brand, so the score reflects discovery, not recall.
3. 03Each prompt is answered live with web search enabled, the way buyers see it.
4. 04We parse every answer for whether you are named, your position, which competitors are named instead, and which sources were cited.
5. 05The result is a 0 to 100 visibility score, a mention rate, share of voice against competitors, and a prioritised fix plan.

Why it differs from SEO rankings

• Answers are built from a handful of trusted sources (review sites, comparison articles, Reddit threads, documentation), not from your homepage alone.
• Being named on G2, in a "best tools" roundup and in Reddit threads matters more than keyword density.
• Answers vary between runs, so visibility should be sampled across many prompts and tracked over time, not checked once.

What moves the number

• Category clarity: your homepage says plainly what you are and who it's for.
• Third-party coverage on the sources AI cites for your category.
• Comparison and alternatives pages that answer the questions buyers ask.
• Structured data, a clean sitemap and an llms.txt so assistants can read you accurately.