Your buyers ask ChatGPT before they ask Google.Find out what it says about you.
No account, no card. The free scan takes a few minutes.
Your brand sits at the centre. We put your buyers' questions to all four engines around it — every line is an answer traced back to who it named and cited.
480 sampled answers, not one screenshot
A single answer from a language model is noise. The whole product is the sample size, and the method behind every figure is printed next to it.
Answers sampled per audit — 40 buyer-intent prompts, 4 engines, 3 passes each. One answer from a language model is noise; 480 is a measurement.
Engines queried natively, with web grounding on. Where they disagree is itself a finding — strong in Perplexity and absent from Gemini is a different problem.
From payment to report. No call, no onboarding, no retainer. Enter a domain, get a prioritized fix list.
Not a score. A shopping list.
A visibility score tells you how bad it is. The audit exists for what comes after: knowing exactly which answers you lose, to whom, and which pages would change that.
See exactly where you're invisible
Presence broken out per engine and per buyer-intent class. “Strong on vendor questions, absent from comparisons” is a strategy; a single blended number is a mood.
Who wins the questions you lose
Every prompt where a rival got named and you didn’t, with the rivals ranked. Losing to a competitor is fixable; not knowing you are is not.
The exact pages to go get
The third-party pages the answers are actually built from — G2 lists, comparison posts, community threads — ranked by citation count, with URLs.
A before/after you can prove
Your prompt set is frozen and replayed verbatim on a re-scan, so the 90-day comparison is real — not two unrelated samples.
Numbers with formulas attached
Every figure states its sample size, its dates and its method. The score weights are published, so you can argue with them.
Domain in, fix list out
No onboarding call and nothing to install. The pipeline is the same one the methodology page documents — this is just the short version.
Enter your domain
No account, no install, no access to anything. We read your public site to infer your category, your likely competitors, and the questions your buyers ask.
We sample 480 answers
40 buyer-intent prompts, put to ChatGPT, Claude, Gemini and Perplexity with web grounding on — three passes each, because one answer from a language model is noise.
You get a fix list
A scored report within 24 hours: where you’re invisible, who wins instead, and a prioritized, effort-tagged list of the pages to go get.
This is what the report looks like
Not a mockup — these two sections are rendered by the shipping report code from demo data, and labelled as such. The sample report shows every section.
| Brand | Claude | ChatGPT | Gemini | Perplexity |
|---|---|---|---|---|
| Acmeanalyticsyou | 36% | 33% | 27% | 48% |
| Clearscope | 88% | 89% | 77% | 99% |
| Semrush | 73% | 82% | 68% | 87% |
| Profound | 46% | 58% | 44% | 75% |
The rows that make it actionable are the rivals': absent from Gemini while your competitor isn't is a different problem from absent everywhere.
- semrush.com45%(217)Claude · ChatGPT · Gemini · Perplexityhttps://semrush.com/
- g2.com36%(172)Perplexity · Claude · ChatGPT · Geminihttps://www.g2.com/categories/seo-tools
- reddit.com36%(171)Claude · Gemini · Perplexity · ChatGPThttps://www.reddit.com/r/SEO/comments/1c8xk2p/best_ai_visibility_tools/
- clearscope.io35%(168)Claude · Perplexity · Gemini · ChatGPThttps://clearscope.io/
- capterra.com33%(160)Claude · ChatGPT · Gemini · Perplexityhttps://www.capterra.com/seo-software/
- news.ycombinator.com33%(158)Claude · Perplexity · ChatGPT · Geminihttps://news.ycombinator.com/item?id=41203881
Your own domain highlighted — or conspicuously missing. This list is what turns “get cited more” into specific URLs to go get.
Four numbers, each with a formula you can check
Not a single unfalsifiable “AI visibility score”. Every metric below is defined in the methodology, computed from the sampled answers, and reported with the spread across passes.
Presence rate
The share of sampled answers that name you at all.
Citation rate
The share that cite your own domain as a source, not just mention you.
Share of voice
Your mentions measured against every rival named alongside you.
Average position
Where in the answer you land — first named, or an afterthought.
Most tools stop at the score
The free scan tells you whether you have a problem. The audit is the part that tells you what to do about it.
Free scan
Enough to know whether you have a problem.
- A visibility score across all four engines
- Presence, citation rate, share of voice, average position
- Your closest competitors, ranked against you
- The top five sources the engines cite
- A prioritized fix list
- Every cited URL, and your frozen prompt set
Five prompts, one pass each. Takes a few minutes.
AI visibility audit
The part that tells you what to do.
- 480 sampled answers instead of 20
- Every cited source domain, ranked, with the URLs
- A prioritized fix list — effort-tagged, tied to specific prompts
- Which of your buyers’ questions you lose, and who wins them
- Full competitor leaderboard, per engine
- Your frozen prompt set, so a 90-day re-scan is a real before/after
One-time. No retainer, no seat, no sales call.
Launching something new? The citation layer is being claimed right now.
When a category is new, the answer engines settle on one canonical source for it — and every later competitor gets explained in that source’s vocabulary, from that source’s pages. The audit shows whether the category terms in your space are open, contested, or already lost, and who is winning them.
Through the APIs, with grounding on — and we publish the tradeoff
We query each provider's API with web grounding enabled — not by scraping the consumer ChatGPT or Google AI Overviews interfaces. That makes the numbers reproducible and the prompt set auditable, and it means they aren't byte-identical to what a person sees in those apps.
We publish that tradeoff instead of hiding it, along with the sample sizes, the engines, the dates, and the exact score formula. If a competitor won't show you theirs, ask why.
And no, there are no testimonials on this page. When a customer quote appears here it will be a real one, with a name attached. A trust section you can fake is not a trust section.
- Print a number we did not measure
- Claim a result from a single answer
- Scrape a consumer interface and call it an API result
- Hide a failed engine run from the denominator
- Invent a testimonial, a logo wall, or a star rating
Before you run the scan
- How is this different from my SEO tool’s new AI module?
- Ahrefs, Semrush and the dedicated monitoring platforms sell ongoing tracking as a subscription, which is the right shape once you know you have a problem and are working on it. This is the step before that: a one-time measurement that tells you whether you have a problem, how bad it is, and what specifically to do — without a procurement conversation.
- Do I need to install anything or give you access?
- No. You enter a domain. We only read your public pages — to infer your category and competitors and generate the prompt set — and everything else happens against the engines’ APIs on our side.
- What exactly arrives for $349?
- A report built from 480 sampled answers: your visibility score with the four components underneath it, per-engine and per-intent breakdowns, the full competitor leaderboard, every cited source domain with URLs, a prioritized fix list tagged by effort, and your frozen prompt set for a re-scan later. One-time payment, no retainer.
- What if the engines never mention my brand at all?
- That is the most common starting point, and the report is built for it. Absence is not a dead end — the answers are usually assembled from a small set of third-party pages, and the report tells you which ones, so “get cited more” becomes a list of specific URLs rather than a slogan.
- Why should I trust the numbers?
- Because you can check them. The methodology is published — engines, sample sizes, prompt construction, the exact score formula and its weights — and every report discloses its own dates, denominators and failed runs. The sample report on this site is generated from demo data and labelled as such; we never print a number we didn’t measure.
Find out what the engines actually say about you.
The free scan runs five prompts across all four engines and shows you the score, your closest competitors and the top cited sources. It takes a few minutes, and it is enough to tell you whether the $349 audit is worth it.