OpenAI
ChatGPT
Responses API with web search; the citations in the answer go into the source list.
SEO and GEO analysis tools
SEO Checker Google Rank Checker Google Rank Tracking Keyword Analysis AI Visibility Checker Schema Checker llms.txt Generator DA PA CheckerCalculation tools
VAT Calculator Discount Calculator Percentage Calculator WhatsApp Link GeneratorWhat ChatGPT, Gemini and Claude answer to the same question: is your brand named?
The AI visibility checker measures whether your brand is named unprompted in ChatGPT, Gemini and Claude answers. Up to three questions about your sector are put to the three assistants with web search, in the same way; the questions do not contain your brand name. You see which brands are recommended for the question your customer would ask.
The result holds a visibility score from 0 to 100, a status for each assistant (named, cited as a source, not named), the competitor brands named most and three recommendations on why the brand is not named. The detailed report, with shortened excerpts from the answers and the cited sources, goes to your e-mail as a PDF.
Visibility in AI search means your brand appears in the answer when someone asks ChatGPT, Gemini or Claude a question. These assistants search the web, choose sources and often put a few brand names forward. A brand that ranks high on Google may not appear in these answers at all.
GEO (Generative Engine Optimization) is the work that gets a brand into these answers: crawler access to the site, structured data, content that answers the question directly and the brand being named on trusted sources. GEO is part of every inurdemirel SEO package; this test is the baseline measurement for that work. The concept in detail on our blog: What is GEO?
OpenAI
Responses API with web search; the citations in the answer go into the source list.
Grounding with Google Search; the pages the answer rests on go into the source list.
Anthropic
Web search tool; the pages quoted in the answer go into the source list.
Each question is put to the three assistants separately: OpenAI's ChatGPT model, Google's Gemini model and Anthropic's Claude model, each with its own web search. The assistants are asked to answer as they would for a regular user and to list the brands they name; your brand name is not in the prompt.
Your brand name or domain is searched for in the answer; differences in case, Turkish characters and joined spelling are tolerated. Whether your site is cited as a source, in which position the brand is named and the other brands in the answer are recorded too. If an assistant does not respond at that moment, the test completes with the other two and the result says so.
Each answer is scored: brand named first 100, second or third 85, later 70; brand name absent but your site cited as a source 35; not present at all 0. The GEO score is the average of all answers received; each assistant's score is the average of its own answers. An answer that could not be retrieved is left out of the score.
Score per answer
| Status | Score |
|---|---|
| Named first | 100 |
| Named second or third | 85 |
| Named later | 70 |
| Cited as a source only | 35 |
| Not named | 0 |
The dark summary card on screen shows the visibility score, the status of the three assistants and the competitors named most; below it come three recommendations and an SEO + GEO package recommendation that fits your situation. The detailed PDF report holds a shortened excerpt of every question's answer from every assistant, the brand's position, the brands named and the cited sources.
The PDF is sent only to your e-mail address; every report has an unguessable number that can be verified on our site. The share link shows the summary only and is valid for seven days. Questions and answers are kept for 30 days and then deleted.
The most common reasons: AI crawlers blocked in robots.txt, a site without structured data that clearly defines the brand and the service, pages that do not answer the question directly, and the brand missing from the guide, list and news sites the assistants cite. The test also measures robots.txt and llms.txt; the recommendations are chosen from these findings and the sources cited in the answers.
The SEO Checker measures the technical and content side with 20 GEO checks; the llms.txt Generator covers crawler access and llms.txt. To be present on the cited sources, the backlink packages fit; for ongoing GEO work, the SEO packages.
GPTBot, OAI-SearchBot, ClaudeBot, Google-Extended and PerplexityBot must be able to read the site.
Question-style headings, the answer in the first 60 words, Organization and FAQPage schema.
The brand name on the guide, list and news sites the assistants cite.
Technical SEO, content, backlinks and GEO work in monthly packages; visibility is measured and reported every month. The package recommendation in the result is the starting point.
From ChatGPT to the GEO score: a short answer to every question.
Measure VisibilityYou enter your brand name, domain, sector and up to three questions your customers might ask. The test puts the same questions to ChatGPT, Gemini and Claude with web search and shows whether your brand is named in the answers, in which position, and which brands are recommended instead.
The test measures whether the brand is named unprompted. If the brand name is in the question, the assistant repeats it anyway and the result does not show real visibility. That is why questions containing the brand or domain are not accepted.
The GEO score summarises the brand's visibility in AI answers from 0 to 100. Each answer is scored by mention and position (first 100, second or third 85, later 70, source only 35, not named 0); the score is the average of all answers.
Yes, no sign-up or card details are asked for. Because every test sends paid requests to three AI providers, there is a daily credit: each test uses 1 credit, with 2 credits a day. For our customer panel members the daily credit limit is 5. The same brand and questions are answered from a 7-day cache without using a credit.
AI answers vary with the wording of the question, the search results at that moment and the model version. The result is a single measurement on the test date. To see the trend, the measurement is repeated with the same questions at regular intervals.
The questions are sent to OpenAI, Google and Anthropic through our server so that they can produce answers; no request goes from your browser to a provider. Questions and answers are kept for 30 days and then deleted. Personal data (name, phone, e-mail, ID number) should not be entered; questions containing such data are not accepted.
The recommendations section of the report starts with three steps: crawler access for AI bots, content and structured data that answer the question directly, and presence on the sites the assistants cite. With a free analysis request our team compares the results with your competitors and proposes an SEO + GEO plan.
Every report carries an unguessable number. When it is checked at our verification endpoint, it returns the domain, date and tool and nothing else. That shows whether a report presented to you was really produced by this tool.
For your questions, project or ideas, our expert team is with you 24/7. The relevant team reviews your request and gets back to you.