Lead Scoring

Heuristic 0–10 score. File: src/enrichment/processors/scoring.ts. Not ML — additive weights based on which signals were extracted.

Source: docs/SYSTEM_OVERVIEW.md § Lead scoring.

Formula

let score = 0;
if (contacts.some(c => c.role_type === 'VD')) score += 2;   // VD found
if (contacts.some(c => c.phone))              score += 1;
if (contacts.some(c => c.email))              score += 1;
if (contacts.length >= 3)                     score += 1;
if (domain)                                   score += 1;
if (google_maps?.rating)                      score += 1;
if (market_intel?.growth_signals?.length)     score += 1;
if (email_patterns_detected?.length)          score += 0.5;
 
return Math.min(Math.round(score * 10) / 10, 10.0);

Maximum

8.5 of 10 from the listed signals. Headroom reserved for future signals (NPS, financials).

Inputs

• contacts[] from Crawlee Scraper / Firecrawl
• role_type from inferRoleType() (VD, Styrelseordförande, CFO, CTO, etc.)
• domain from Domain Discovery
• google_maps from Google Places
• email_patterns_detected (e.g. firstname.lastname@)

Caveat

Note

Heuristic, not validated against revenue or conversion. Works for POC ranking, not for ML-grade qualification.

See also

EnrichV7, Crawlee Scraper, Google Places, Domain Discovery.