How we measure AI-search visibility
A score is only useful if you can trust how it's built. Here's exactly how GenAI Ranker measures whether AI assistants recommend your brand — and how we tie it to real revenue.
1. Real buyer questions, not keywords
We build a bank of the actual high-intent questions your buyers ask AI assistants ("best running shoes for flat feet", "X vs Y", "alternatives to Z"), tagged by funnel stage — awareness, consideration, conversion. You track the prompts that decide purchases.
2. Asked across all 6 engines
Each prompt is put to ChatGPT, Gemini, Claude, Perplexity, Grok and Google AI Overviews. We capture the verbatim answer, whether your brand is named, its position, the sentiment, the competitors named, and the sources cited.
3. A transparent 0–100 visibility score
Each answer is scored on whether you're mentioned, how prominently (position), and how favorably (sentiment) — then averaged into a single visibility score per engine and overall. Higher means AI names you more often, earlier, and more positively. No black box.
4. Refreshed daily — and cached for efficiency
Paid brands are re-measured every day, so you catch movement fast. Identical prompt × engine × day results are cached, keeping monitoring efficient without sacrificing freshness.
5. Attribution: from visibility to revenue
A lightweight pixel (or our WordPress/Shopify apps) detects visits referred by AI assistants and ties conversions and revenue back to them — so you see the business outcome, not just the score.
6. Benchmarks, anonymized
We compare your score to your category and the wider platform using aggregated, anonymized data with a minimum sample size — never exposing any individual brand.
7. Citation Authority — the sources AI trusts
Every answer records the domains AI cited. Aggregated across the whole network and keyed only on a topic's keywords (never a brand or prompt), that reveals the source domains AI trusts most for your category — and which you're not cited on yet. It's gated by k-anonymity (a source only shows once enough distinct brands share the topic), so it's a privacy-safe target list, not a leak.
8. The Autonomous Agent — safe by default
Opt in per brand and the agent picks your highest-value gap from your own recommendations (or a Citation Authority target), drafts the fix, and either queues it for your approval — the default — or, only if you explicitly turn it on, auto-publishes to your connected site. Every cycle is capped monthly and logged, so it never runs up cost or changes a live site without your say-so.
What's a “tracked prompt”?
One buyer-intent question we monitor for your brand across the engines. Your plan's allowance is the total tracked prompts across all your brands — the unit that actually maps to buying decisions, so you pay for coverage that matters.
See your score — measured this way, free
Type your brand and run it through the full methodology at no cost.