Two kinds of points
Separate who the lead is from what the lead has done. Mixing them hides good-fit leads that are simply quiet.
- Fit points — industry match (+20), company size in range (+15), decision-maker title (+15), in your service area (+10).
- Engagement points — asked for a quote (+25), replied to an email (+10), visited the pricing or services page twice (+5).
- Negative points — student or job-seeker (−30), competitor (−50), free email domain for B2B (−5).
Pick a threshold and a time window
Decide the score at which a person must follow up — for example 50 — and how fast. Engagement points should decay: a page visit from three months ago is not a buying signal today.
Check it against reality every quarter
Take the deals you won and lost last quarter and look at their scores. If won deals often scored low, your fit criteria are wrong. If high scores rarely close, you are rewarding activity rather than intent. Adjust the weights, not the threshold, first.
Common mistakes
Most scoring models fail for the same few reasons.
- Too many rules — start with six to eight and add only when the data shows a gap.
- No negative scoring — junk leads float to the top.
- No owner — someone must be responsible for reviewing the model.