The Lokrix AI Score

The Lokrix AI Score is a single, calibrated number from 0 to 100 that summarises how visible your property is across the AI answer engines — always reported with a 95% confidence interval, never as a bare point estimate. It rolls up per-engine component scores, presence, share-of-voice and citations into one figure you can track over time.

What the score measures

The score answers one question: for the prompt universe your buyers actually ask, how likely is an answer engine to surface your brand — by citing, linking, recommending or naming it? Because engine answers are non-deterministic, Lokrix runs each prompt many times and aggregates the outcomes into a calibrated probability rather than a single yes/no. The headline score is a weighted composite of those probabilities across every prompt and every engine in scope.

Anatomy of the Lokrix AI Score gaugeA 270-degree radial gauge with a hairline track, a blue value arc drawn from the start to the score, a translucent confidence-interval band, nine evenly spaced minor ticks, a value marker, and a colored band-health dot; a legend labels each element.7263–80Value arc — the score, 0–10095% confidence band (lo–hi)Nine minor ticks (0, 12.5, … 100)Value marker at the scoreBand-health dot (green/amber/red)

Calibrated, not a vanity metric

“Calibrated” means the number reflects real probability: when Lokrix reports a 70% presence likelihood, roughly 70% of comparable runs surface your brand. The confidence interval widens when evidence is thin — few samples, cached data, or a narrow prompt set — and tightens as you run more live samples. Treat the interval as part of the score, not decoration.

How to read it

  • The number — your overall 0–100 visibility, weighted across engines and prompts.
  • The interval — the 95% CI. A wide band means “sample more before you trust the trend.”
  • Provenance — every contributing data point is labelled live, cached or mock (sample). Mock data is never presented as live.

Go deeper

Break the score apart in Score anatomy, compare engines in the per-engine breakdown, and understand the underlying presence, SoV & citation metrics. To see how the samples behind the score are produced, read sampling & confidence intervals.