Facts & warnings

What your scores can and can't tell you.

Compare scores only on the same setup

Warning: scores only compare when the questions, the engines and the location are the same.
  • Different questions measure different things, even for the same site on the same day.
  • Each engine answers differently, so turning engines on or off changes the score.
  • To measure progress, turn on tracking (every run keeps the same questions, engines and location) or duplicate a scan.

Your questions shape your score

  • Questions that don't match how customers search score near zero, like a local business's questions without its city.
  • Questions with your name or terms only you use score high but tell you little.

Answers vary between runs

  • A few points up or down can be noise. Watch the trend over several runs.
  • Tracked runs ask again when your visibility on an engine flips, and keep the majority answer.
  • Engines update their models. If every engine moves at once, that's a likely cause.

Location

  • Google, Bing, DuckDuckGo, Yahoo, ChatGPT and Claude get the scan's location. Gemini, Perplexity and Grok always answer globally.
  • The location doesn't translate your questions.

How the score works

  • Named in an answer: up to 60 points, less the lower you're listed.
  • Linked as a source: up to 40 points.
  • Overall: half the average answer score, half whether most engines name and link you on each question, weighted by estimated search volume.

Fact checks and tone

  • Only answers that name you are checked.
  • A wrong fact is reported only when text on your own website contradicts it. What your site doesn't say can't be checked.
  • An AI model judges tone and facts, so read the answer before acting on a flag.
  • Included in the paid plans.

Competitors

  • Suggested automatically and may include businesses that aren't real competitors. Add your own, or keep only the ones you list, when you set up a scan.