Why doesn't ChatGPT mention my business?

Because nothing ChatGPT can read gives it a reason to. When an assistant answers “who is good at X?”, it works from what it learned in training and, increasingly, from live sources it retrieves. If your site never plainly says what you do and no source it trusts mentions you, the honest output is silence. It is rarely personal and almost never a penalty. It is absence of evidence.

The cause is usually one of four, and they are worth checking in order because each has a different fix.

1. Your site never answers the questions being asked

Models assemble recommendations from pages that address the question. If your site says “reimagining wellness journeys” instead of “physiotherapy clinic in Leeds treating running injuries”, a model cannot confidently place you in any category. The fix is to publish direct answers: one page per real buyer question, question as the heading, answer in the first paragraph.

2. Machines cannot read the site

Some sites are effectively invisible to AI crawlers: all content rendered by JavaScript, text baked into images, or a robots.txt that blocks GPTBot, ClaudeBot or PerplexityBot outright. Check the robots file first, it takes a minute. Then look at your site with JavaScript off; if the page is empty, so is some crawlers’ view of you.

3. Nobody the model trusts mentions you

Assistants lean on third-party evidence: comparison articles, review platforms, directories, community threads. A business with zero footprint outside its own domain gives the model nothing to corroborate. Find the specific pages that get cited when your category questions are asked and get present on them, or publish the better version of that page yourself.

4. The answer slots are already owned

For some questions a handful of well-established names dominate every answer. You will not displace them head-on quickly. The route in is specificity: narrower questions, locations, niches and comparisons where the current answers are thin. Win those first; broader visibility follows the evidence trail you build.

Finding out which one applies to you

Guessing between these four wastes months. Measuring takes minutes: ask the assistants your buyer questions, note where you are absent, read which sources their answers cite, and check your site’s machine-readability. We built Sourcive to run exactly that diagnosis: it measures your visibility across ChatGPT, Claude, Perplexity and Gemini, shows who is being recommended instead of you and why, and produces the specific pages and files that fix it.