Skip to main content
The API computes every figure with the same code as the dashboard, so a number from the API equals the number on screen for the same site, range and filters.

​
Which answers count

  • Only successfully collected answers in the range.
  • A Google result page that showed no AI Overview is left out of every rate. It says something about the question, not about your brand.
  • Competitor figures use confirmed competitors only. Unconfirmed brands found in answers are measured but not ranked until someone confirms them in the dashboard.

​
Why past numbers can change

Figures are calculated when you ask, from the stored answers. Which brands an answer named, and which of its links were your own site, are fixed when the answer is collected, so a new alias or domain counts from then on. Ranks and competitor figures for a past range change if you confirm, archive or edit a competitor, and category groupings change if you move a prompt. See Why past numbers change.

​
Compared with Profound’s API

If you are moving from Profound, note:
  • Profound’s visibility_score is a 0–1 share of answers that mention any brand. RankRaft’s is the 0–100 composite above, the same as our dashboard.
  • Profound files days in US Eastern time; RankRaft uses UTC for answer reports.
  • Profound averages each metric per engine and then across engines. RankRaft pools all answers in the range; use group_by: ["platform"] for per-engine figures.