What the fit score means

The score is not a single opinion from a model. The AI produces structured signals — ICP match, role relevance, company fit, buying signals, timing, data confidence — and the final number is computed from those in code. That is why the breakdown always adds up.

Data confidence is the part people miss. A profile with a headline and nothing else scores differently from one with ten years of history, because there is less to judge. A low score on a thin profile often means 'not enough to say' rather than 'not a fit'.

The score answers one question: how well does this person match the ideal customer profile you described in onboarding. If the scores feel wrong, the profile is usually the thing to fix — Settings → AI sales profile.