Key Takeaways
- Read the shape before the total. A 125-check result is 25 questions asked of five assistants. Two businesses named 15 times each can be in very different positions, depending on which questions and which assistants those 15 came from.
- Look first at the questions closest to hiring. Being named on specific questions, the ones a buyer asks when they are nearly ready, is worth more than the same count on broad ones.
- Count agreement across assistants. A question where three of five assistants name you is a sturdier finding than one where a single assistant did. Answers vary from run to run; agreement survives that.
- Check what each answer said about you. A mention that describes you wrongly, or files you under the wrong specialty, is not a win. Read the answers, not just the tally.
- Compare against yourself, not an average. There is no industry benchmark for a 125-check result. The honest comparison is your own next run, on the same questions.
A 125-question result arrives as a number, and the first instinct is to judge the number. Fifteen sounds low. Forty sounds good. Neither reading tells you much on its own.
What good looks like in a 125-question result is a pattern, not a total: named on the questions closest to hiring, by more than one assistant, in words that describe you correctly. Two businesses with the same count can be in very different positions.
Below is what a 125-check result actually contains, two illustrative results with the same total and opposite shapes, five checks to read your own honestly, and what even a good result does not mean.
What does a good 125-question result look like?
A good 125-question result names you on your most specific buyer questions, has several assistants agreeing on those same questions, and describes you accurately when it does. The count is the last thing to judge, not the first.
If your result comes from the 25-question free check instead, the bands in what a good AI visibility score looks like are the better guide. This post is about the larger grid, where the extra dimension, five assistants instead of one, changes how to read it.
What a 125-check result actually is
A 125-check result is 25 buyer questions each asked of five assistants: ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews. It is a grid, 5 rows by 25 columns, and every cell is one answer that either named you or did not.
Each cell is one reading, not a permanent fact. A study published by SparkToro with Gumshoe in January 2026, in which 600 volunteers ran 12 prompts through ChatGPT, Claude and Google's AI a combined 2,961 times, found less than a 1 in 100 chance of getting the same list of brands twice. It measured product lists rather than local services and is not peer reviewed, but the direction is clear: any single cell can flip.
The assistants also build answers differently. Google's documentation on AI features and your website describes answers assembled from multiple related searches, and each assistant searches its own way. That is why the rows of the grid rarely look alike, and why a pattern across rows is worth more than any one of them.
Same total, different shape
Two results with the same total can mean opposite things. The figure draws two illustrative 125-check results, questions ordered from broadest on the left to most specific on the right. Both businesses are named 15 times.
Business ANamed 15 of 125, all on broad questions, 9 by one assistant
Business BNamed 15 of 125, all on five specific questions, 3 assistants each
Same count. A has no question named by more than two assistants. B has five questions, 17 to 21, each named by three.
Both businesses are illustrative, not clients. Business A is named 15 times, all on the broad half of the questions, nine of them by Perplexity alone and none by Claude. No single question names it more than twice. That is a thin, fragile position: most of it rests on one assistant and on questions a buyer asks early, when they are still browsing.
Business B is also named 15 times, but every mention sits on five specific questions, 17 to 21, and each of those five is named by three of the five assistants. That is a strong position for a small practice. It owns the questions buyers ask when they are close to choosing, and the finding is sturdy because the assistants agree.
Five honest checks, in order
Read a 125-check result with five checks in a fixed order: which questions, how much agreement, what the answers said, who was named alongside you, and only then the total. The order stops the total from coloring everything after it.
| Check | What good looks like | The common misreading |
|---|---|---|
| 1. Which questions | Named on the specific questions closest to hiring | Counting broad mentions as equal to specific ones |
| 2. Agreement | Two or more assistants name you on the same question | Treating one assistant's mention as settled |
| 3. What was said | Correct name, city, specialty and offer | Counting a wrong description as a win |
| 4. Who else was named | Peers of your size, not only national brands | Ignoring the rival list, which is the real finding |
| 5. The total | Higher than your own last run on the same questions | Comparing it with someone else's number |
The fourth check deserves more time than it usually gets. The names that appear in your place tell you who the assistants consider your alternatives, and what those businesses have that you do not. How to work through that list is in how to read an AI visibility report.
What a good result does not mean
A good result does not mean you will stay named, that every buyer sees you, or that there is nothing left to fix. It is one reading of a moving system, and it describes answers, not bookings.
Even Business B above can slip if a competitor publishes a better page for question 19, or if one assistant changes how it searches. That is why the useful comparison is always your own next run on the same questions. How far a result can reasonably move, and over what time, is covered in how to set a realistic 90-day AI visibility target.
The damaging admission: a small practice rarely looks like Business B on its first run, and a first result that looks like Business A is common and not a failure. It tells you which questions you are closest to winning. And if you are already named on your core questions by most assistants, you may not need anyone's help; the honest reading is to keep doing what you are doing.
If you have not run anything yet, start with the free check. It asks ChatGPT 25 real buyer questions from your niche at $0 and emails a plain-English readout: your score, who got named in your place, and the one fix to start with. It runs as soon as you ask for it, and the report lands within the hour between 7am and 5pm Central, or by 10am the next morning outside those hours.
Questions we hear the most
What does a good 125-question AI visibility result look like?
A good result names you on the specific questions closest to hiring, with several assistants agreeing on the same questions, and describes you accurately. The total matters less than where the mentions sit.
Is 15 out of 125 a good score?
It depends on the shape. Fifteen mentions spread thin across broad questions from mostly one assistant is a weak position. Fifteen concentrated on your core questions, with three assistants agreeing on each, is a strong one for a small practice.
Why does agreement between assistants matter?
Because a single answer is a sample. AI answers vary from run to run, so a mention from one assistant may not repeat. When several assistants name you for the same question, the finding is much more likely to hold.
Should I worry if one assistant never names me?
Not on its own. Assistants draw on different sources, and a gap on one is common. It becomes worth looking at when that assistant is the one your buyers use most, or when it names the same competitor every time.
Is there an average 125-question score to compare against?
No. The result depends entirely on which questions were asked and which niche and city they cover. Compare a result with your own earlier and later runs on the same questions, never with someone else's number.
What does the AnswerHalo $500 audit show beyond the total?
Every answer. The audit runs 125 checks across ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews, keeps each answer so you can read it, and shows who was named instead of you and which sources the answers leaned on. It lands by this time tomorrow.
Is the free check the same as the 125-question audit?
No. The free check asks ChatGPT alone 25 real buyer questions from your niche at $0, and emails your score, who got named in your place and one fix, within the hour between 7am and 5pm Central, or by 10am the next morning.
Your buyers are already asking. Find out what AI tells them.
The free check asks ChatGPT 25 real buyer questions about your niche and sends you the report within the hour, at $0. If it shows you are already getting named everywhere, we will say so in plain words.
Get my free check →