Skip to main content
GEO Confidence Score is a measure of how statistically reliable a brand’s AI visibility metrics are, based on sample size (number of prompts and runs) and response variance (how much results differ across runs). A high confidence score means the reported mention rate and other metrics are stable and trustworthy; a low confidence score means more data is needed before drawing conclusions.

​
Why it matters

AI visibility data is inherently noisy due to non-determinism. A brand with 25 prompts run once each has less reliable metrics than a brand with 75 prompts run multiple times. Without a confidence indicator, users might over-react to a mention rate drop that is actually just statistical noise. Confidence scoring helps users know when to act and when to wait for more data.

​
How Cited uses it

Cited computes confidence from prompt count, run count per prompt, and cross-run variance (standard deviation of mention positions). The confidence_level field in the daily metrics data categorizes each measurement as high, moderate, or volatile. Dashboard users can filter to see only high-confidence metrics when making strategic decisions.