On this page

1. What Mention rate counts
2. Why it comes first
3. Read it with its likely range
4. Tell a real change from ordinary variation
5. What else moves the figure
6. Engine by engine
7. How to raise your Mention rate

When a buyer asks an assistant which products to consider, the first thing to know is whether your name is in the answer. Mention rate counts exactly that, across every question you track and every engine you ask.

What Mention rate counts

Mention rate is the share of answers that named you. An answer here is one question answered by one engine, so ten questions on six engines give 60 answers a week. If 15 of those 60 named you, your Mention rate is a quarter.

A few details decide what goes into the count:

• Only questions that leave your name out. A question that names you gets an answer that names you almost by definition, so BotSay leaves those out of the five measures. What remains are the mentions you earned.
• Only answers that came back. A question that was declined, hit an error or got no Google AI Overview is left out of the count. It is never counted as though you went unmentioned.
• Repeats are averaged. Where a question was asked more than once in a week, BotSay averages within the question first, then across questions.

The engines are pooled into one figure. An engine that answered more of your questions counts for more in it.

Why it comes first

The other four measures read your answers in different ways, and two of them depend on Mention rate. Average rank says how early you were named, and Sentiment says how warmly you were described, but both count only answers that named you. A fine Average rank on a handful of answers can sit beside a low Mention rate. Share of voice counts names rather than answers, so it rises and falls with your rivals' mentions too.

That is why BotSay puts Mention rate first on This week, with the headline beside it saying whether it moved. When you report one number about AI answers, this is the one to report.

Read it with its likely range

A Mention rate counted from a few answers can swing on a single reply. Counted from a few hundred, it settles. So BotSay shows Mention rate with two things beside it: the number of answers it was counted from, and its likely range, the span the true figure probably sits in.

The range comes from the count of answers. More answers give a narrower range, and fewer give a wider one. It is worked out at the standard confidence level used in statistics, not by asking a question again to see how it wobbles. Mention rate is the one measure with a likely range, because it is a share of the answers read and a range has a clear meaning for it.

The number of questions you track sets how narrow the range can get. Our guide How many questions do you need to track AI visibility? shows the range at each plan's size.

Tell a real change from ordinary variation

Answers vary from one week to the next even when nothing about you has changed. BotSay calls a change in Mention rate real only once it is bigger than the uncertainty in both weeks' readings put together. On a steady week, the reading says so and tells you how big a change would have needed to be to count, so you know where the bar sits next week.

The picture below shows eight weeks of Mention rate for an invented brand, each week with its likely range. Seven weeks hold steady, then the latest week rises clear of the one before.

On Standing, the Over time tab draws the same kind of chart for your own readings, with the range as a band behind your line. Where two weeks' bands stay apart, the weeks certainly differ. The full method behind the headline can also call a change real when the bands overlap a little. How we measure sets out the test.

One timing detail helps here. ChatGPT, Perplexity, Gemini and Claude answer on your reading day, and Google AI Overviews and Google AI Mode answer the day after. Until the week is complete, BotSay marks the figure Provisional and gives no verdict, so a half-read week never reads as a fall.

What else moves the figure

Your own visibility is the change you are tracking. A few other things can move the number too.

• Your questions. Adding or stopping questions changes which answers are counted. Change your list in batches, and note the date.
• Your names. You count as named under the names listed for your brand in Setup. If buyers and engines use a short name, an old name or a product name, add it, or answers that name you that way will not count.
• What Google shows. Google decides for itself whether to show an overview for a search. When it shows fewer, the rate is worked out over the questions it did answer.

Engine by engine

Each engine answers in its own way, so your Mention rate can differ a good deal from one to the next. The picture below shows an invented brand's Mention rate on each of the six engines, with the rate across all six at the top.

In BotSay, the By engine tab on Standing shows your figures on each engine and marks the strongest and the weakest. Its last row blends all six into the figure This week shows. When one engine names you clearly less than the rest, a card counts the questions where it named a rival and not you, with a Read button that opens that engine's answers on Evidence. The card shows you where to look, and the answers show what the engine leans on.

Use the figure across all six for your headline. Each engine's figure rests on a sixth of the answers, so its range is wider, and it is best read over a few weeks.

How to raise your Mention rate

Mention rate rises when you start being named in answers that name your rivals today. Start with the questions where that happens most.

1. Find the questions you are missing. The Questions page shows how many answers named you on each question and which rivals were named instead. The Never chip lists the questions where you have not been named yet.
2. Read those answers. Look at what the engine says about the rivals it named, and which pages it linked to.
3. Write the page or earn the listing. Where a rival's own page is cited, write the page that answers the same question for your product. Where review sites, communities or publishers are cited, get listed or take part there. Our post Pages AI assistants cite goes through both.
4. Watch the questions you gained and lost. The Gained and lost tab lists questions that started or stopped naming you. A single answer flipping is often chance, so look for the same question changing on several engines.

BotSay's Work list does the ranking for you: topics where a rival is named ahead of you, ordered by how many questions you lose on each, with the questions and the pages they cite. When a page goes live, mark it published, and Is it working compares how often answers named you on that piece's questions before and after.

Keep reading

Measures

AI visibility results: how to tell if a page worked

How to compare AI answers before and after a page goes live, read the result with confidence, and know what a steady reading tells you.

6 Nov 2026 · 6 min readRead →

Measures

Sentiment in AI answers: how engines describe you

What Sentiment measures in AI answers, how to read the quotes behind it, and the pages and listings to fix when an engine describes you poorly.

30 Oct 2026 · 6 min readRead →

Measures

Average rank in AI answers: how to move up the list

What Average rank counts in AI answers, why it can slip in a good week, how to read it beside Mention rate, and the work that moves you up the list.

23 Oct 2026 · 6 min readRead →