What is lead scoring?
How does explainable lead scoring work?
With rule-based lead scoring, the company defines which response earns how many points: a landing page login more than an opened email, an appointment request more than a click. Points can decay over time so that interest from three months ago does not carry the same weight as interest from yesterday. The result is transparent: you can say at any time why a contact has 72 points.
Predictive methods, by contrast, estimate the probability of closing using statistical or AI models. They can find additional patterns but are harder to explain. In regulated industries, it has therefore proven effective to start with a rule-based model and use predictions only as a supplement.
The score is used to set priorities: above a threshold, the journey notifies an adviser, a contact moves to a different segment or receives a more specific offer. Because lead scoring evaluates personal aspects, it is generally profiling within the meaning of the GDPR – transparency towards customers and the right to object to direct marketing, including related profiling (Art. 21(2) GDPR), must be taken into account.
Marlene Hoffmann logs in to her landing page with two-factor authentication and spends 2:40 minutes reading the “tax advantage” argument. Her score rises to 72. Because the threshold is 60, adviser Christina Reuter receives a task: offer an appointment. (Example values)
Distinction
| Term | Difference |
|---|---|
| Lead nurturing | Communication process controlled by the score |
| AI prediction (predictive scoring) | Estimates probability using a model rather than fixed points |
| Credit scoring | Assesses creditworthiness; a different purpose with its own legal requirements |
How the PBM Campaign Platform supports it
Rule-based, explainable points model, configurable and with time decay, on two levels (campaign score and contact score); optionally an AI prediction layer that can be switched off.
Related pages
-
Platform
Analytics & reporting -
Use case
Cross- & upselling