What Does a Good AI Visibility Score Look Like for a Small Coaching Business?

Somebody ran a check on your business and handed you a number. Now you want to know whether it is good, and there is no table anywhere that tells you.

That is not because the information is being withheld. It is because a visibility score is not a standardized measure. It depends on which questions were asked, in whose words, across which assistants, on what day. Two checks of the same business can differ by a lot and both be honest.

So this post does the useful version instead: what the number counts, what each range actually means for a small practice, and the cases where a low one is the correct result and you should leave it alone.

What does a good AI visibility score look like?

For an independent coach or consultant, being named in about four to eight of twenty-five buyer questions is a strong position. Nought is normal at the start. One to three is where most small practices sit once they have any public presence at all. Nine or more usually means a national brand, a directory, or a franchise rather than a person.

Those bands are ours, from our own runs, and they describe shape rather than a benchmark you can hold anyone to. We publish them because the alternative is a client staring at a number with nothing to compare it against, and inventing a worse comparison in their head.

The reason the ceiling is so much lower than people expect comes down to arithmetic. An answer has room for roughly three names. Twenty-five questions span everything from a broad category search to a narrow local one, and no independent practice is a sensible answer to all of them.

What the number is actually counting

A count of answers that named you. Twenty-five questions asked, some number of answers came back with your business in them, that number is the score. That is the entire calculation, and it is deliberately something you could reproduce by hand in an afternoon.

We say that plainly because the alternative is common and worth being suspicious of. A proprietary index that blends mentions, sentiment, and position into a single figure out of a hundred cannot be checked by the person paying for it. A number you cannot recreate is a number you have to take on trust, and nothing here is shown on trust.

What the count does not capture is worth naming too. It does not know whether you were mentioned first or third, whether the description was flattering, or whether the question was one your buyers ask. Those are judgment calls that belong next to the number, not inside it. Google's guidance on AI features in Search is a useful reminder of why: answers are composed fresh each time, so a count is a sample of behavior, not a rank.

Four bands, and where a small practice realistically sits

Here is the whole range with what each part of it means in practice.

MENTION COUNT ACROSS TWENTY-FIVE BUYER QUESTIONSILLUSTRATIVE

Absent Named in none of the twenty-five Nothing to measure yet. The first job is existing in text somewhere other than your own site.

Occasional Named in one to three You appear on the narrow questions and vanish on the broad ones. This is where most small practices start.

Established Named in four to eight A realistic target for an independent coach inside a year. You hold a slot in your specialty.

Dominant Named in nine or more Usually a national brand, a directory, or a franchise. Rarely the right goal for a solo practice.

The strip runs from absent on the left to dominant on the right. Segment widths show roughly how much of the range each band covers, not how many businesses land in it. Bands are illustrative and drawn from our own runs.

AnswerHalo answerhalo.com

CLICK TO ENLARGE

Written out, because nothing here may live only inside a figure. Absent means named in none of the twenty-five, and the first job there is existing in text somewhere other than your own site. Occasional means named in one to three, where you appear on narrow questions and vanish on broad ones, and it is where most small practices start. Established means named in four to eight, a realistic target inside a year, holding a slot in your specialty. Dominant means named in nine or more and usually describes a national brand, a directory, or a franchise. All four bands are illustrative.

Notice which band we are not telling you to aim for. Chasing the fourth band as a solo practitioner means competing on broad category questions against organizations with hundreds of pages written about them, and that is a bad use of a year, which is the same conclusion reached in do you actually need help with AI visibility.

Why the questions matter more than the score

A score of two can be better than a score of six. If those two mentions came on the questions your buyers actually type and the six came on questions nobody asks, the two are worth more, and no single number will ever tell you that.

That last row is the failure mode to watch for in anybody's report, including ours. It is trivially easy to produce a flattering score by choosing questions a business already wins. The check is to read the twenty-five questions and ask whether a real buyer would type them, which is the discipline behind the questions to ask ChatGPT about your own business.

When a low score is fine and you should ignore it

If your practice is full and every client arrives by introduction, a low score is a fact about the public web, not a problem with your business. It becomes a problem when the referral pipeline thins, or when the people being referred start checking you before replying. Until then it is information, not an emergency.

If your buyers are not asking assistants yet, the same applies. Some markets have moved and some have not, and the way to find out is to ask your last five clients how they checked you before the first call rather than to trust a vendor's urgency, ours included.

And if you are brand new, a nought is simply accurate. There is nothing published about you because nothing has happened yet. The honest advice is to go and do the work that gets written about, then measure.

What does deserve attention regardless of the number is the direction. Run the same twenty-five questions again in a month with the wording unchanged and compare the lists rather than the totals. That comparison costs nothing and is more informative than any single score, and the method for keeping it is in how to track AI visibility without buying software.

One last note on why the count moves at all. Princeton researchers testing generative engine optimization found that citing sources and quoting authority raised how often a page was used in generated answers, and Perplexity's help center describes answering with sources attached. Both point the same way: the number goes up when more quotable text about you exists in more places, and not for any other reason.

Our free check runs twenty-five buyer questions from your niche at $0 and emails a plain-English readout within the hour, including the count and the questions behind it. It runs as soon as you ask for it.