Lead Scoring Model
The prompt
Design a lead scoring model.
What we sell: {{describe}}
ICP: {{firmographics_and_persona}}
Data we have: {{fields_collected_title_company_size_page}}
CRM/marketing tool: {{what_will_implement_scoring}}
Please design:
1. Demographic fit scoring (firmographic attributes)
2. Behavioural scoring (actions and their point values)
3. Negative scoring (signals that reduce score)
4. MQL threshold: at what score is a lead marketing-qualified?
5. Score decay: how should scores decrease if a lead goes inactive? Why this works
Separating demographic fit scoring from behavioural scoring reflects the two dimensions that actually predict sales-readiness — who someone is and what they've done. Including negative scoring explicitly prevents the model from ignoring disqualification signals like competitor job titles or unsubscribes. Asking for a handoff threshold forces the team to pre-define the MQL/SQL boundary rather than debating it deal-by-deal.
Risks & review
Lead scoring models are only as good as the data fields you actually collect and trust. The AI will design a logical scoring framework but point values require calibration against your historical win data — the first version will be a hypothesis, not a validated model. Plan to review and adjust the scoring after 90 days of operation based on whether MQL-to-SQL conversion rates improve.