Buyers ask an assistant before they search. We put your category question to ChatGPT twice, in the same minute, and show you both answers: whether you were named, who was named instead, and which sources the model actually drew on.
We ran one prompt three times, minutes apart, with everything held constant. Sixteen sources appeared across the three answers and four appeared in all of them.
The top of the ranking. The same two brands took first and second place in every run, in the same order. If you are already the obvious answer in your category, you will find that out quickly and it will keep being true.
That part is stable enough that measuring it once is fine.
Everything else. The fourth-placed brand appeared in two runs out of three. Three quarters of the sources changed between runs, including the competitors' own documentation.
That is the part you would be paying to understand, and it is the part a single run reports as a confident fact.