Combine firmographic fit with a small number of meaningful behaviours, add negative signals and validate the model against accepted opportunities.
Separate fit from engagement
A perfect-fit account with no activity is different from an active student or supplier. Keep the dimensions visible so sales can understand why a lead is prioritised.
Reward high-intent behaviour
A pricing request, use-case visit or return to implementation content usually means more than several blog views. Weight actions according to the decision they indicate.
Use negative scores
Personal email addresses, irrelevant countries, careers-page activity and repeated low-value downloads can lower priority. Negative scoring protects the team from inflated totals.
Close the feedback loop
Review which scored leads sales accepted, rejected and converted. Adjust the model using observed outcomes rather than leaving it untouched after implementation.
A practical checklist
- Define ideal-customer fit separately
- Choose five to eight high-intent actions
- Add negative criteria
- Show sales the reasons behind the score
- Review outcomes every month
Frequently asked questions
What score should trigger sales follow-up?
There is no universal number. Set the threshold using historical leads and sales capacity.
Should email opens receive points?
Usually very few, because opens are noisy and do not necessarily show buying intent.
Can lead scoring work with low volume?
Yes, but a simple rules-based model is often more reliable than a complex predictive model.