Free tool · GEO Visibility Checker
AI visibility checker: do ChatGPT, Perplexity, Gemini and Claude recommend you?
Find out whether AI assistants recommend you — or send your buyers to a competitor.
In short
The GEO Visibility Checker measures whether your brand is mentioned or cited in sampled AI answers to buyer questions on the platforms you select, scored 0–100. You get a visibility score with evidence of where you are named or cited per platform, your share of voice against competitors, and the next steps to earn your place in the answer.
This check is temporarily unavailable
The GEO Visibility Checker only ever shows live data, so the form is switched off until the check is back. Ask us to run it for you — free, by hand, with the same evidence the tool would give.
This check is temporarily offline
The GEO Visibility Checker cannot return live data right now, and it never shows estimated or made-up results. Ask us to run it for you — free, by hand, with the same evidence the tool would give.
Request a free manual checkWhat this checks
Coverage: The free check queries ChatGPT, Perplexity, Gemini and Claude — the four AI platforms with web-search APIs — and reports only on the platforms it actually queried. Google AI Overviews, AI Mode, Copilot and other assistants are outside this free check; ask us about covering them in a paid audit.
- Whether each platform names your brand in its answer
- Whether your website appears among the sources it cites
- Which of your competitors are mentioned instead, and whether you are named before or after them
- A 0–100 visibility score from mention rate, citation rate, position and share of voice — overall and per platform
- The answer excerpt and source list for every question, per platform
Limitations
- AI answers vary with wording, time, location, account history and model updates; one prompt is a sample, not a measurement of universal visibility.
- Only the platforms listed as queried were checked, using API access with web search enabled — consumer apps may answer differently. Requests that failed are shown as such, never dropped.
- Mentions are detected by a whole-word match of the brand name or website host in the answer text; citations by matching the website host in the sources the platform returned (a source whose title is the bare host counts too). Nothing is inferred or scored by a model.
- The visibility score is our own documented formula (mentions 40, citations 30, position 15, share of voice 15). It summarises this sample; it is not a ChatGPT, Google or Perplexity metric.
- Some platforms return redirect or proxy URLs as sources; where the real host is not visible, a citation can be missed.
What you get
- A visibility score with evidence of where you are named or cited per platform, your share of voice against competitors, and the next steps to earn your place in the answer.
- Free and without sign-up: up to 5 checks per visitor per day; identical inputs within a few hours reuse the stored result.
- Every result names its data source and the time of the check, lists its limitations and comes with a shareable report link.
- Answers vary with wording, time and location; one sample is not a universal ranking, and only queried platforms count.
How to use it
- Enter your brand name and website URL. Mentions are matched on the brand name, citations on the website host (www is ignored and subdomains count).
- Describe what you sell in a few words, as a customer would say it, choose the market and optionally your industry — they form the first line of every prompt.
- Type one to five customer questions, one per line, exactly as a buyer would ask them. Add up to five competitors to see who is named instead of you.
- Select the AI platforms to query and run the check. Every selected platform receives every question with web search enabled; most checks take 15–90 seconds.
- Leave your email only if you want us to follow up. The shareable report link works without it.
How to read the results
- The 0–100 visibility score summarises this sample: 75 and above reads as strong, 50–74 moderate, 25–49 low, below 25 barely visible. It is our documented formula, not a platform metric.
- The question × platform matrix shows, for every answer, whether your brand was mentioned, whether your website was cited and the order in which the names appeared.
- Per-platform cards give mention and citation counts and the model that answered; the competitor table shows share of voice — your mentions divided by all brand and competitor mentions.
- Every answer excerpt (the first 400 characters) and its source list are included, so you can verify each count yourself.
- Recommendations are rule-based and derived from the gaps found — no citations, competitors named first, uneven platforms — a starting list, not a strategy.
- The report is stored at a shareable link. Identical inputs within six hours reuse the stored result and are marked “reused recent check”.
Methodology
How the check works
- Each question is wrapped in one fixed prompt that names your market, product and industry and asks the platform to recommend companies by name and include the sources it relied on.
- Detection is mechanical: a mention is a case-insensitive whole-word match of your brand name or website host in the answer text; a citation is a returned source URL whose host matches your website; competitors are tested the same way. No model judges anything.
- Position is 1 plus the number of entered competitors named before you (1 = named first).
- Platforms are queried through their APIs with web search enabled — one request per platform per question, three in flight at a time, 25-second timeout — so consumer apps may answer differently.
I'm a customer in {market} looking for {what you sell} in the {industry} space. {your customer question} Please recommend specific companies, brands or websites by name, explain briefly why, and include the source links you relied on.
Words in braces are filled from the form; the industry clause is dropped when you leave that field empty. One prompt is sent per platform per question, with web search enabled.
Visibility score weights
| Component | Weight | Value (0–1) |
|---|---|---|
| Mention rate | 40 | answers naming the brand (or its website host) ÷ answers |
| Citation rate | 30 | answers citing the website as a source ÷ answers |
| Position | 15 | mean of 1 ÷ position over all answers (1st = 1, 2nd = 0.5, …; not mentioned = 0) |
| Share of voice | 15 | brand mentions ÷ (brand + competitor mentions) |
Score = Σ weight × value ÷ Σ weights used, rounded. Without competitors, share of voice is left out and the other weights are scaled up, so the score stays on a 0–100 scale. 75 and above reads as strong, 50–74 moderate, 25–49 low, below 25 barely visible.
What is GEO?
GEO is the practice of making a brand and its content more likely to be mentioned and cited in answers produced by generative AI systems such as ChatGPT, Gemini, Claude and Google AI Overviews.
Unlike classic ranking, a generative answer is composed from several sources, so GEO works on being a clear, consistent, verifiable source: unambiguous entity information, well-structured pages, third-party corroboration and content that directly answers buyer questions. Results are observed by asking the platforms real questions and recording mentions and citations, which vary by wording, time and location.
FAQ
GEO Visibility Checker questions
No. It is our own documented formula — mention rate 40, citation rate 30, position 15, share of voice 15 — applied to the answers that came back. It summarises one sample on one day; the platforms publish no visibility score of their own.
Let's get on a call
Book a free consultation meeting below, and let’s understand what you actually require
- Free audit, no obligation
- Reports from your own Search Console & GA4
- No fabricated metrics — ever