What Is Lead Scoring? Models & Examples
Also known as Lead score, Predictive lead scoring, Lead qualification scoring
Lead scoring is a method of ranking leads by how likely they are to become customers, usually by assigning points for fit and engagement. Fit points reflect how closely a lead matches your ideal customer profile, such as company size and job title. Engagement points reflect actions like requesting a demo. High scores go to sales first.
Lead Scoring explained
Most models combine two dimensions. Fit, sometimes called explicit or demographic scoring, uses firmographic and role data: industry, headcount, seniority, department, tech stack. Engagement, or behavioral scoring, uses actions: email clicks, page views, webinar attendance, content downloads, product usage in a trial. Many teams also subtract points for negative signals, like personal email addresses, students, competitors or long inactivity.
Scores feed routing and lifecycle stages. A lead might become a marketing qualified lead (MQL) once it crosses a threshold, then be passed to sales for follow-up. Predictive scoring uses historical win data and machine learning to weight factors automatically, but it needs enough closed deals and clean CRM data to work well.
A simple, transparent model that sales trusts beats a complex one nobody understands. Keep fit and engagement visible as separate numbers, so reps can tell a perfect-fit account that has not engaged yet from an eager lead at a company too small to buy. Add score decay so old activity fades, and review the model with sales every quarter.
Lead Scoring example
A B2B software company scores leads out of 100. Fit: +20 for 50 to 500 employees, +15 for a target industry, +15 for director level or above. Engagement: +20 for a demo request, +10 for a pricing page visit, +5 per webinar. Minus 30 for a personal email domain. Leads above 60 go to sales within an hour; those between 30 and 60 enter a nurture sequence.
How to build a lead scoring model
Analyze won and lost deals
Compare the traits and actions of leads that became customers against those that did not.
Define fit criteria
Assign points for firmographic and role traits that match your ICP, and subtract points for clear disqualifiers.
Define engagement criteria
Give more points to high-intent actions like demo requests and pricing page visits than to light actions like opens.
Set thresholds with sales
Agree on the score that triggers sales follow-up and what response time sales commits to.
Review and recalibrate
Every quarter, check whether high-scoring leads actually convert better and adjust points that are not predictive.
Common mistakes
- Giving points for email opens, which are unreliable because of privacy features and bot activity.
- Combining fit and engagement into one number so reps cannot see why a lead scored high.
- Building the model without sales input, so reps ignore the scores.
- Never adding score decay, so a lead that engaged a year ago still looks hot.
- Setting and forgetting the model instead of checking it against real conversion data.
Lead Scoring FAQ
There is no universal number. A good threshold is one where leads above it convert to opportunities noticeably more often than leads below it. Many teams start with a score out of 100 and set the sales handoff point after reviewing a few months of conversion data.
Rules-based scoring uses points that people assign manually based on experience, such as plus 20 for a demo request. Predictive scoring uses historical data and machine learning to find which factors correlate with conversion and weights them automatically. Predictive needs enough clean data on won and lost deals.
A lead often becomes an MQL when its score crosses a set threshold, showing enough fit and engagement for marketing to hand it over. Sales then reviews it and, after qualifying it through conversation, may convert it to an SQL. Scoring is the mechanism behind the MQL stage.
Yes, in a different form. Outbound teams often score accounts and contacts before outreach using fit and signals like funding, hiring or intent. This helps decide which accounts get personal research and which go into lighter sequences. Replies and meetings then feed back into the score.