Lead Scoring
Assigning a numeric value to a lead that predicts conversion likelihood — historically rule-based, increasingly driven by ML models trained on past conversions.
Quick answer
What is Lead Scoring?
Assigning a numeric value to a lead that predicts conversion likelihood — historically rule-based, increasingly driven by ML models trained on past conversions.
Last updated:
Lead scoring tries to answer: "Of all the leads in my pipeline, which should I work first?" Traditional scoring assigns points: job title matches ICP (+10), company size > 500 (+5), opened 3 emails (+3). The total is the score.
ML-based scoring (which Salesforce's Einstein Lead Scoring and gptfy's ML Recommendation Engine implement) replaces the rule book with a model trained on past conversions — the model learns which combinations of attributes actually predict closed-won, weighted automatically.
Key implementation considerations: training data quality (need at least 12 months of closed/lost outcomes), feature engineering (which attributes to consider), and explainability (sales managers want to know *why* a lead scored high). Modern AI-powered scoring includes natural-language explanations alongside the score, addressing the black-box concern.
Related terms
Browse all terms- Lead RoutingThe process of automatically assigning incoming leads to the right sales rep based on rules (territory, industry, deal size) or AI-driven matching.
- Sales ForecastingPredicting future revenue from pipeline data, historical close rates, and (increasingly) AI signals — used for capacity planning, reporting, and goal-setting.
See it in your Salesforce org
See Lead Scoring running in GPTfy
Book 30 minutes with a GPTfy engineer to see how Lead Scoring actually works inside a Salesforce org like yours.
Book a demo