Lead Scoring

Formula: Lead Score = sum of weighted actions and attributes, with negative points subtracted for bad-fit signals. Example: visiting the pricing page is worth 10 points, downloading a guide is worth 5, opening a cold email is worth 1. A free-tier email domain or a company size outside your ICP subtracts 5. A lead crossing 20 points becomes an MQL worth sales' attention. Two kinds of signals feed the score: behavioral (what someone does, like visiting a page or attending a demo) and firmographic or demographic (who they are, like company size or job title, checked against your ICP). Behavioral signals show intent; fit signals show whether that intent is worth anything to you. For small teams: don't build an elaborate scoring model before you have the lead volume to need one. A spreadsheet tracking three signals that someone actually checks beats an automated scoring workflow in HubSpot or Customer.io that nobody looks at. Add complexity only once manual triage becomes the bottleneck. Revisit the model regularly. Scores set once and never checked against actual close rates drift, a signal that predicted purchases a year ago may not predict them now.

A scoring model gives 10 points for a demo request, 5 for downloading a case study, and subtracts 10 for a personal email domain when the product is B2B-only. A lead who requests a demo with a work email scores 10, clears the 15-point MQL threshold once they also open a follow-up email, and gets routed to sales same day.