Skip to main content

What Knownwise measures

What a run asks, which assistants answer, and what every answer is judged for.

Last checked against the product on October 10, 2026

These docs are in English.

Knownwise tracks how often AI assistants mention, rank and recommend your brand when they answer the questions your buyers ask.

What a run is

A run asks each of your tracked questions to each assistant: ChatGPT, Claude, Gemini and Perplexity. On the Free plan, runs use ChatGPT and Perplexity.

What every answer is judged for

  • Mentions — whether the answer names your brand or one of your competitors.
  • Position — where your brand stands in the answer, when the answer ranks it.
  • Recommendation — whether the answer picks your brand as the one to choose.
  • Sentiment — whether the mention reads positive, neutral, mixed or negative.
  • Citations — which sources the answer cites, and whether they are yours.

An assistant may answer a question without searching the web. Those answers count like any other. Only citations come solely from answers that searched.

How answers are counted

One answer is one question asked to one assistant. Each plan sets how many answers it allows, and a run spends one answer per question per assistant. When the allowance is spent, scheduled runs pause.

The Free plan's allowance covers your first report only. You read the results on your brand overview.