Lead Scoring Model Review Prompt
Prompt
You are a marketing/revenue operations manager reviewing the lead scoring model. Lead scoring data: [PASTE: Scoring criteria (demographic + behavioral) | Points per criterion | Current MQL threshold score | MQL-to-SQL conversion rate | MQL-to-close rate | Rep feedback on lead quality | Any criteria not in the model that reps say they'd find valuable] Review the model for: 1. Predictive validity — do high-scoring leads actually convert at higher rates than low-scoring? 2. Recency weighting — are recent behaviors weighted more heavily than old ones? 3. Negative scoring — are unqualified signals (wrong company size/industry, bounced email) reducing scores? 4. Threshold accuracy — is the MQL threshold set at the score that predicts genuine sales-readiness? 5. Missing signals — behaviors or attributes that reps find predictive but aren't in the model Output: Lead scoring model assessment. Changes to improve predictive accuracy. Recommended threshold adjustment. Expected improvement in MQL quality.
Why it works
Reviewing the model against actual MQL-to-close rates rather than just MQL-to-SQL rates validates whether scoring is identifying customers who ultimately buy, not just customers who engage with marketing content. Comparing lead quality by score tier identifies whether the scoring thresholds are appropriately calibrated — an MQL that converts to SQL at 30% but closes at 2% suggests the scoring is identifying engaged-but-not-buying leads. Including rep feedback captures the qualitative intelligence about lead quality that quantitative data alone misses.
Watch out for
Lead scoring model reviews that result in score changes must be implemented gradually — a sudden change in the MQL threshold will affect pipeline reporting, sales team expectations, and marketing attribution in ways that need to be managed as a change programme, not a technical update. Communicate scoring changes to both sales and marketing leadership before implementing, and restate prior period metrics if the threshold change significantly affects historical MQL volume.
Used by