What the Flare score is
The Flare score is Brandflare's headline 0–100 measure of how AI engines present a brand. It is deliberately not a single metric: it combines five separate measurements of the same set of judged answers into one trendable number, and always shows the breakdown behind it.
It replaces mention rate as the default ranking on Explore and on every brand dashboard. Mention rate itself survives as Flare's largest component.
The five pillars
| Pillar | Weight | Question it answers |
|---|---|---|
| Presence | 30% | Do engines name you at all? |
| Prominence | 22% | How high and how forcefully? |
| Accuracy | 22% | Is what they say true? |
| Reach | 16% | Is it consistent across all four engines? |
| Sentiment | 10% | Is the framing favourable? |
The full formulas — including the rank decay curve, fair-share scaling for share of voice, and severity weighting for accuracy flags — are in the Flare score methodology.
Why not just use mention rate
Mention rate rewards being talked about, and nothing else. Two brands can post identical mention rates while being in entirely different positions: one named first and described correctly on all four engines, the other named last on one engine with a false claim attached. A single-metric leaderboard cannot tell them apart, which is the whole reason Flare exists.
Two properties worth knowing
It is absolute, not a percentile. A brand's Flare moves only when its own measurements move — never because a competitor improved. That is what makes the trend line mean something, and it is why the score is not a rank.
Pillars abstain rather than guess. When a measurement was never taken — accuracy is not assessed for lite-tier catalog brands, for instance — that pillar drops out and the remaining weights re-normalise. A brand is never awarded points for a check nobody ran, and the on-screen breakdown always shows the weights actually used.
Reading it
Direction first, then the breakdown. A score falling from 78 to 71 matters less than which pillar fell: losing seven points to Accuracy is a fact-sheet and PR problem, while losing them to Reach means one engine has stopped surfacing you and the others haven't. The composite tells you something changed; the pillars tell you where to look.