Mention-Rate Confidence Interval (Wilson)

Enter how many AI answers named your brand out of how many, and get the mention rate with a range you can defend.

How to use

What it checks and how it works

The mention rate is the answers that named the brand divided by all answers. The range uses the Wilson score interval, which behaves well with small samples and with rates near 0% or 100%, where the simple normal (Wald) interval can run below 0, above 1 or collapse to a point. The table shows both so you can see the difference. The normal critical value comes from the confidence level you choose. The interval assumes every answer is an independent draw. Repeated runs of the same few questions are correlated, so the true uncertainty is wider than shown: spread your answers over many questions. It says how sure you can be about the rate in the answers you collected, not that the rate holds for every question a buyer might ask.

Why Wilson instead of the simple plus-or-minus formula?

The simple formula can give a range that falls below 0% or above 100%, and it is unreliable when the sample is small or the rate is near an extreme. The Wilson interval stays between 0 and 1 and keeps its coverage better in those cases, as the statistics literature on binomial intervals describes.

What does a 95% interval mean?

If you repeated the whole data collection many times and built an interval each time, about 95 in 100 of those intervals would contain the true rate. It is a statement about the method, not a 95% chance for this one interval.

My brand was named in none of the answers. What does the tool say?

It gives a rate of 0% and an upper limit above zero, which shows how high the true rate could still be given the number of answers you checked. With few answers that upper limit can be large.

Sources

Related tools

Privacy

Runs in your browser; nothing is uploaded. The interactive tool needs JavaScript.