The essentials
- Definition: lead scoring assigns points to each contact according to their profile and their behaviour, so leads can be ranked by likelihood of purchase.
- Two dimensions: fit with your ideal customer (sector, size, job title) and observed engagement (pages viewed, content downloaded, emails opened, quote requested).
- How to use it: triggering the move to MQL, prioritising callbacks, adapting sequences, and removing off-target contacts from campaigns.
- The rule: start with ten criteria at most, a single threshold, and refine every quarter from the deals actually signed.
The aim of Lead scoring is the method of assigning points to each contact in a database, so they can be ranked by likelihood of purchase and sales time concentrated on the most promising. A score of 75 out of 100 triggers an immediate callback; a score of 20 leaves the contact in an automated email sequence. It is the mechanism that gives operational substance to the distinction between MQL and SQL : rather than a case-by-case judgement, a numerical rule decides the handover to sales.
The two dimensions to rate
| Dimension | What it measures | Example criteria |
|---|---|---|
| Fit (profile) | The match with your ideal customer | Sector, headcount, revenue, the contact's job title, geographic area, technologies used |
| Engagement (behaviour) | The intent being shown | Pricing pages viewed, quote requested, demo booked, decision-stage content downloaded, emails clicked, replies to follow-ups |
The two dimensions are rated separately, which distinguishes four situations: good fit and strong engagement (call now), good fit and weak engagement (nurture with content), poor fit and strong engagement (often a student, a competitor or a job applicant), poor fit and weak engagement (remove from the active database).
A starter grid for a small business
| Criterion | Points |
|---|---|
| A target sector | + 15 |
| Headcount between 20 and 250 | + 15 |
| A decision-making role (director, marketing manager) | + 15 |
| A business email address | + 5 |
| A quote or demo request | + 30 |
| A visit to the pricing page | + 15 |
| A decision-stage download (comparison, client case study) | + 10 |
| Three or more visits in 30 days | + 10 |
| Opens with no clicks for 60 days | minus 10 |
| No activity for 90 days | minus 20 |
| Competitor, student, job applicant, supplier | Excluded |
With this grid, a threshold of 60 points triggers the move to MQL and the sales callback. The negative points matter as much as the positive ones: without a decay mechanism, an entire database eventually crosses the threshold within two years.
Setting up scoring
- Start from your signed customers: list the last twenty deals won and identify what they had in common before signature; those commonalities are your criteria, not your intuition.
- Limit yourself to ten criteria: a simple model that sales understands beats a clever one nobody applies.
- Set a single threshold: one MQL handover to begin with; multiple thresholds come later.
- Automate it in the CRM: the score calculates and updates itself, from your site and email data; connecting the sources is described in connecting your marketing data.
- Measure how predictive it is: each quarter, compare the score at the moment of handover with the deal's actual outcome; if leads at 80 points don't sign more often than those at 50, the grid is wrong.
- Revise: adjust the weightings from the deals won and lost, together with sales.
The frequent mistakes
- Rating engagement alone: a devoted blog reader who isn't in your target will never sign; profile must weigh as much as behaviour.
- An over-complex model: forty weighted criteria nobody can explain produce scores sales ignores.
- No decay: with no negative points and no expiry, everyone ends up qualified.
- Scoring without volume: below 30 leads a month, prioritising by hand works better.
- Never checking it: a grid never tested against signed deals degrades within a year, all the more so as your target shifts.
- Confusing score with value: the score measures likelihood of purchase, not size; a large account at 55 points sometimes deserves more attention than a small one at 80. The value metrics are covered in CAC: definition.
How GreenRed helps
Rather than juggling several tools, the GreenRed action plan brings these metrics together in a single dashboard, compares them over time and tells you which actions come first. You can try it free, with no card, from the Pricing.
Frequently asked questions
What is lead scoring?
A method that assigns points to each contact according to their profile (sector, size, job title) and their behaviour (pages viewed, quote requested, emails clicked), so leads can be ranked by likelihood of purchase and sales time concentrated on the most promising.
Which criteria should you use to score a lead?
Two families, rated separately: fit with your ideal customer (sector, headcount, job title, area) and engagement (a quote or demo request, a visit to the pricing page, decision-stage content, visit frequency). Negative points penalise inactivity and exclude off-target profiles.
At what volume does lead scoring become useful?
From around thirty leads a month. Below that, prioritising by hand works better, and the model lacks the data to be verified. Beyond a hundred leads a month, it becomes hard to do without.
How do I know whether my scoring grid is any good?
By comparing, each quarter, the score at the moment of handover to MQL with the deal's actual outcome. If your best-scored leads don't sign more often than the rest, the grid predicts nothing and its weightings need reworking from the deals won.