Key Takeaways
- Test in fresh sessions. Your own account history can skew answers toward you. Log out or use a temporary chat so you see what a stranger sees.
- Use buyer words, not your words. Ask "who should I hire to fix my nutrition", not "best certified holistic practitioner". The test only works in their language.
- Count names, not feelings. Write down every business named in every answer. A tally is evidence; an impression is a mood.
- Repeat before you conclude. AI answers vary between runs. Ask each question twice and look for the pattern, not a single verdict.
- Act on the biggest gap. One measured weakness fixed well beats five guesses. Score first, then fix the reason that costs you the most answers.
Here is how to check if AI recommends you, in one sitting, with nothing but a browser: ask five buyer questions in fresh sessions on two AI tools, write down every business named, and count how often you appear. Fifteen minutes, a real number at the end.
Most coaches and consultants have never run this test. They have opinions about AI, worries about AI, sometimes a screenshot a friend sent. What they do not have is a tally, and the tally is the only part that can be acted on.
This guide is the manual version of what we run at scale. It is genuinely enough to know where you stand, and we will say so plainly at the end, including the cases where you should not hire anyone.
How to check if AI recommends you: the 15-minute version
The test: write five questions a buyer would ask, run each one on ChatGPT and Perplexity in fresh sessions, and log every business named. Score yourself out of ten answers. That is the entire method.
Two tools matter because they gather evidence differently. ChatGPT leans on what it learned in training plus web results when it searches, as OpenAI's explainer describes. Perplexity builds each answer from live pages and shows them, as its help center documents. A gap that shows up on both is a real gap.
One rule before you start: fresh sessions only. These tools can remember your past chats, and asking about your own niche while logged in can name you more often than a stranger would ever see. Use a temporary chat, a private window, or a logged-out session.
Minutes 1 to 5: write the five questions buyers actually ask
Write questions in the buyer's words, not yours. Buyers do not ask for a "certified transformational practitioner". They ask who is best, who to hire, whether a program is worth it, and what the alternatives are.
Use this pattern and fill in your niche:
- "Who is the best [your niche] coach for [your ideal client]?"
- "Who should I hire to help me [the outcome you sell]?"
- "Is [your name or program] legit? What do people say?"
- "What are the best alternatives to [the big name in your niche]?"
- "Is hiring a [your niche] coach worth the money?"
The third question matters even if you expect silence. An answer that knows nothing about you is a different problem, with a different fix, than an answer that describes you wrongly.
Minutes 6 to 12: run the questions and log the names
Work through the questions methodically and write everything down as you go. Memory smooths results; a log does not. Here is the schedule, minute by minute.
| Step | Minutes | What you do | What you write down |
|---|---|---|---|
| 1. Write the questions | 1 to 5 | Draft five buyer questions using the patterns above | The five questions, exactly as you will ask them |
| 2. Open fresh sessions | 6 | Temporary chat on ChatGPT, private window on Perplexity | Nothing; just do not skip it |
| 3. Ask on ChatGPT | 7 to 9 | Run all five questions, one per message | Every business named, per question |
| 4. Ask on Perplexity | 10 to 12 | Run the same five questions | Every business named, plus the sites it lists as sources |
| 5. Score it | 13 to 15 | Count answers naming you, and the most-named competitor | Your score out of 10, their score out of 10 |
On Perplexity, the source list is a bonus finding. Those are the pages the tool trusted enough to build an answer from, which means they are the places where being mentioned would change your result.
Minutes 13 to 15: score what you see
Your score is simple: of the ten answers, how many named you? Count the most-named competitor the same way. The gap between those two numbers is your visibility gap, in plain sight.
In that illustrative example, the coach appears in 2 of 10 answers while the top competitor appears in 7 of 10, and one answer names nobody. Do not over-read any single row. The pattern across ten answers is the finding; one answer is an anecdote.
What your score means, and what to do next
Named in 7 or more of 10: you are winning, and you do not need to hire anyone. Keep publishing what you publish, re-run this test quarterly, and spend your money elsewhere. We tell people this in free check readouts and mean it.
Named in 3 to 6: you exist to the machines but lose more than you win. Read the seven reasons AI skips coaches and match your log against them. Usually one gap, often the about page or a missing results page, explains most of the misses.
Named in 0 to 2: you are invisible where buyers are asking, and the fixes are foundational. Start with how your offer is described, then where your proof lives. The shift that makes this worth fixing at all is covered in how clients find a coach now.
A note on tooling: if all you want is this number tracked monthly, a monitoring dashboard subscription can do that, and for some businesses that is enough. The score was never the hard part, though. The fixing is, and that is the work this whole discipline exists to do.
If you want the bigger version of your tally without the manual work, the free check runs 25 buyer questions on ChatGPT and emails the readout within the hour, for $0. It opens soon; the list runs it first. Your five-question score and our 25-question score almost always agree. Ours just comes with the fix attached.
Questions we hear the most
How do I check if ChatGPT recommends my business?
Ask it five questions your buyers would ask, in fresh chats, and write down every business it names. If you appear in fewer than half of the answers, you have a visibility gap worth fixing. The whole test takes about 15 minutes.
Why do I get different answers each time I ask?
Because these tools generate answers fresh each time, results vary by session, wording, location, and model version. That is normal and expected. Ask each question twice and read the pattern across runs instead of trusting any single answer.
Should I test on more than one AI tool?
Yes, use at least two. ChatGPT and Perplexity make a good pair because they gather evidence differently, and a gap that shows up on both is a real gap. A full picture covers five tools, which is what a paid audit is for.
Will my own ChatGPT account skew the results?
It can. These tools can draw on your past chats, so asking about your own niche while logged in may name you more often than a stranger would see. Use a temporary chat or a logged-out session for honest results.
What does AnswerHalo’s free check add beyond this test?
Scale and a readout. Once the check opens, we ask ChatGPT 25 buyer questions from your niche instead of five, then email you a plain-English readout within the hour: your score as one number, who gets named in your place, and the one fix to start with. It costs $0.
How is the $500 audit different from the free check?
The audit runs 125 checks: your 25 questions across all five AI tools, scored and stored. You get the full board, the competitors per question, and six fixes ranked by revenue impact, in your inbox by this time tomorrow.
What happens after I submit the free check form?
Once the check opens, we run your 25 ChatGPT questions and email the readout within the hour. There is no call, no login, and nothing to install. If the readout shows you are already getting named everywhere, it will say so plainly.
Your buyers are already asking. Find out what AI tells them.
The free check opens soon. It asks ChatGPT 25 real buyer questions about your niche and sends you the readout within the hour, at $0. Join the list and you run yours first; if it shows you are already getting named everywhere, we will say so in plain words.
Get on the list →