ICP scoring evaluates each account against the rubric's criteria, applies weights, and produces a score (typically 0–100 or A/B/C tier). The score drives routing, sequence selection, and agent decisions.
The math
Each criterion contributes points based on match × weight. If size criterion is weight 3 and the account matches, it contributes 3 points. Total points are summed and normalized to a 0–100 scale. Disqualifiers override: any disqualifier match forces score to 0.
Tier thresholds
Most rubrics produce A/B/C tiers from raw scores. Defaults: A = top 20% (score 80+), B = next 30% (score 50–79), C = bottom 50% (score below 50). Adjust thresholds based on your motion — if you have AE capacity for only 100 A-tier accounts, set the A threshold to whatever produces 100.
What the score predicts
The score predicts fit, not engagement. A high-score account is one that matches your customer base profile; whether they're actively in-market is a separate question (covered by signals). A complete picture combines fit (ICP score) with engagement (signals + first-party activity).
When scores re-calculate
Continuously, as new data arrives. When an account's firmographics update (new headcount data, industry reclassification, new technographic detection), the score re-runs. Manual rubric changes trigger full re-scoring of all accounts.
Auditing the score
Open any account → ICP Score panel → see the score, the contributing criteria, the weights, and the resulting tier. Useful when an account's score surprises you — the breakdown shows exactly why.
