Frequently asked question · Audit AI Answer Accuracy
Why read a window of answers instead of asking each engine once?
Because one answer is one draw. Ask the same engine the same question twice and you can get a different list of brands, or a different price cited to a different source. A screenshot from one person’s account tells you what that person saw that day. The audit reads every recorded answer for the window, usually the first weeks of tracking, so each claim is logged with how often it appeared and on which days. A wrong claim that turns up in most runs goes to the top of the fix list; one that appeared once gets noted and watched.
What varies between two runs
The same prompt, on the same engine, on two consecutive days, can come back with a different set of brands, a different ordering, and a different source cited for the same fact. The engines retrieve live pages, and the pages they pick change. One study cited on this page found less than a one percent chance that two answers to the same brand question list the same brands. A screenshot is a fair record of what one person saw on one account on one day. It is not evidence about the next person.
How the window is built
The question set runs once a day on ChatGPT, Gemini, Perplexity, Copilot, and Google AI Mode, per market and per persona, and every answer is kept in full. The audit window is usually the first weeks of tracking, which gives the reader a run of answers per prompt per engine instead of one. Every claim is then logged with the engine, the market, the prompt, the cited source, how many runs it appeared in, and on which days.
What the frequency changes
Frequency is what orders the fix list. A wrong price that shows up in most runs on two engines is costing you every day, so it goes to the top. A claim that appeared once in one market is logged, marked, and watched in the next window rather than chased. The saved report carries deltas against the previous window of the same length, so when a fix lands you can follow that claim's frequency and see whether it fell, rather than take our word that it changed.
Answered on
Audit AI Answer Accuracy
What ChatGPT, Gemini, Perplexity, Copilot, and Google AI Mode say about your brand, per market, read claim by claim by a person and traced to its sources.
More questions about Audit AI Answer Accuracy
See what AI is already saying about your brand.
A free audit shows you exactly where you stand across ChatGPT, Gemini, Perplexity, Copilot, Claude, Grok, and AI Overviews, and what it would take to close the gap.