Recovery Scores

The Recovery Score is a transparent, deterministic 0–100 heuristic that estimates how worthwhile it is to re-engage a stale lead.

How it works

Scoring starts from a neutral base and applies a set of weighted factors. Each factor contributes a positive or negative score_effect, and the sum is clamped to 0–100. Because the model is deterministic, the same input always produces the same score.

Positive signals

• An estimate or proposal was sent
• High potential deal value
• No explicit rejection
• Recent inquiry or prior engagement
• Few prior follow-up attempts
• Previous customer or repeat-service opportunity
• High-intent lead source (e.g. referral, website)

Negative signals

• Explicit rejection or opt-out
• Very old lead
• Excessive previous contact attempts
• Invalid contact information
• Closed/lost with a strong negative reason
• Suspected duplicate

Hard ceilings

Some disqualifying signals cap the maximum score regardless of positive factors. For example, an explicit_rejection or opted_out lead cannot score in the high range.

Priority buckets

• high — score ≥ 70
• medium — score 40–69
• low — score < 40

Industry-aware weighting

Each industry profile nudges specific factors. Roofing emphasizes unclosed high-value estimates, HVAC rewards service-renewal timing, and automotive weights unsold internet/showroom leads. These adjustments appear as their own entries in the factors array so the score stays fully explainable.

Reading the factors array

Every response includes a factors array sorted by absolute impact. Each entry names the factor, its impact (positive/negative), and the exact score_effect applied — so you can always show your team why a lead scored the way it did.