AI visibility tests

What we test

For each company, we maintain a rolling set of buyer-intent prompts derived from your category, audience, and product description. Examples in a B2B SaaS category might include:

• "Best [tool] for [audience]"
• "[Competitor] vs [Competitor]"
• "How to [job-to-be-done] without [common pain]"

Where we test

Prompts get sent to six engines in parallel: ChatGPT, Claude, Perplexity, Gemini, and two more in rotation. Each response is parsed for citations and brand mentions, then ranked.

How rank is computed

• Rank 1: you are the primary answer or first cited brand.
• Rank 2–5: you appear in the cited set.
• Rank 6+: you're listed but not cited.
• Unranked: you're not in the response at all.

How often it runs

Visibility tests run once per scheduled audit cadence (weekly on most plans). The Full Domination tier supports daily and on-demand re-runs.

Why rank changes between runs

AI engines are non-deterministic. We send each prompt 3 times and use the modal rank to smooth out variance — but a 1–2 rank wobble between runs is normal even with no changes on your end.