CRM lead scoring assigns points to a lead so the team can see which enquiries look more ready for a salesperson. Points usually come from two ideas: fit (does this enquiry match customers you can serve) and engagement (what has the person done). A threshold then marks a lead as worth a call, worth nurturing, or worth closing out.
Scoring is a model you design. It is not proof that the lead will buy. Many CRM products can calculate a score; that does not mean every deployed CRM already has a useful one. Treat scoring as a configuration decision, and check what your product actually calculates before you promise it to the sales floor.
Fit versus engagement
Fit uses attributes: industry, location, company size, product interest, or whether they asked for a service you sell. A factory in Johor asking for a maintenance contract can score higher for a Johor industrial supplier than a student asking for a brochure.
Engagement uses behaviour: form submitted, pricing page viewed, reply to an email, meeting booked, quotation requested. Opening one marketing email is a weak signal. Asking for a quotation is a strong one.
Keep the two scores separate if you can. A perfect-fit lead who has never responded should not look identical to a poor-fit lead who clicked everything.
A simple points example
| Signal | Points | Why |
|---|---|---|
| Industry matches a served segment | +20 | Fit |
| Located in a region you cover | +10 | Fit |
| Asked for a quotation or demo | +30 | Strong intent |
| Replied to a salesperson | +15 | Engagement |
| Job title is student or personal email only | −15 | Weak commercial fit |
| No activity for 45 days | −10 | Decay |
A working threshold might be: 40 or above, assign to sales within one business day; 20–39, keep in a nurture queue; below 20, disqualify or leave unassigned. The numbers are an illustration. Replace them with your own after you review a month of closed deals.
Where scoring goes wrong
- Too many rules. If nobody can explain the score, people ignore it.
- No decay. A lead that downloaded a file two years ago stays “hot.”
- Vanity engagement. Page views from your own staff, or from a shared office IP, inflate scores.
- One model for every product. A spare-part enquiry and a project tender do not deserve the same weights.
- Score used as a stage. The score suggests priority. The pipeline stage still describes the deal.
Recalculate after you know which leads actually won. If high scores lose and low scores win, the model is backwards.
Putting a score into daily use
- Choose a few fit and intent signals
- Assign points and a decay rule
- Set a threshold for sales follow-up
- Hand off to assignment rules
- Review wins and losses monthly
What to do when the CRM cannot score
You can still rank work. A required “priority” field with three values, set by the person who qualifies the lead, is more honest than a fake automated score. Assignment can then use that field. See how CRM lead assignment rules work.
Key takeaways
- Score fit and engagement separately when possible.
- A threshold should change who acts, and how fast.
- Decay old points.
- Review the model against real outcomes, and do not confuse score with pipeline stage.
FAQ
Is a higher score always a better customer?
No. It means “more like the pattern you coded.” A strategic account can sit at a low score if the pattern ignores relationship deals.
Should we score accounts as well as leads?
Yes, once you sell to existing customers. Lead score helps first response. Account score, or a segment, helps who receives a campaign. Segmentation is covered in how customer segmentation works in a CRM.
How many signals do we need?
Start with five to eight that you can actually capture. Extra signals that nobody fills in only add noise.