Does ChatGPT recommend your business?

How do you find out if ChatGPT recommends your business?

You measure it the way the engines are used: by asking. AI assistants answer customer questions with a short list of named businesses and a set of cited sources. Whether your business appears in those answers is a checkable fact, not a guess — you ask the engines real customer questions and count how often you are named and how often your website is cited.

The mistake most owners make is asking one engine one question — usually "do you know my business?" — and stopping there. Assistants answer differently between runs, the four major engines lean on different sources, and the question that brings you customers is rarely about your name. It is "who should I hire for X" or "what does Y cost near me." A real check covers the questions that carry buying intent, across all four engines.

How do you check your AI visibility in five steps?

1. List 10 questions your customers ask before they buy. Use the phrasing a stranger would type, not your own marketing language — comparison questions, cost questions, "who is the best" questions, and problem questions.
2. Ask each question in ChatGPT, Claude, Gemini, and Perplexity. The engines use different indexes and different sources; being visible in one says nothing about the other three.
3. Record two things per answer: names and citations. Which businesses does the answer name, and which websites does it cite as sources? These are separate outcomes — a business can be named without its site ever being read.
4. Count your share against competitors. How many of the 40 answers name you? How many name each competitor? Which sites get cited instead of yours?
5. Repeat the same questions monthly. Answers vary between runs, so a single check is a snapshot. The trend across identical repeated checks is the real measurement.

This is exactly the structure of a professional AI visibility check; running it manually costs you an afternoon. Tenva runs the same check programmatically across all four engines and walks you through your results free.

What does a real AI visibility check look like?

Here is one from June 2026, run with the method above scaled to 20 questions. The measured business was a software company; its category has well-known competitors. The result is typical of a first check: total absence, while two category leaders dominated the comparison questions.

The same check produced a second finding that matters more than the zero. Absence is common, but so are open answers: on many questions, the engines had no strong source to cite for anyone.

What should you do when ChatGPT recommends competitors but never mentions your business?

Treat it as a diagnosis, not a verdict. The engines are not choosing against you — they recommend what they can find, verify, and quote, and right now that is your competitors. The check tells you precisely which questions you are losing and which sources the engines trust for them. That list of cited sources is your competition for each answer, and it is often beatable: in the June 2026 check above, most cited sources on the open questions were small blogs, not major brands.

The response has three parts. Build a page for each losing question that answers it in plain, quotable language — one question per page, with specifics an engine can lift. Keep your business details identical across your site, directories, and review platforms so the engines can verify you. And start with the open questions, where no strong source exists yet: an unclaimed answer slot goes to the first business that supplies a quotable answer, while a slot held by a strong incumbent takes far longer to win.

What does a free AI visibility checker measure?

A useful checker measures four things: citations (how often each engine cites your website as a source), mentions (how often your business is named in answer text), share of voice (your counts against named competitors on the same questions), and the per-engine split (ChatGPT, Claude, Gemini, and Perplexity behave differently). Citations and mentions diverge more than most people expect — in the June 2026 check, one tool was named in 16 answers but its website was cited in only 1. Mentions reflect reputation the engines absorbed from elsewhere; citations mean your own pages are being read and quoted.

Anything that reports a single score without showing the underlying questions, answers, and cited sources is a marketing widget, not a measurement. The questions are the unit that matters, because each one is a slot your next customer's answer comes from. Tenva's free check runs your business through the full four-engine measurement and shows you every answer.