The Scoring Model Was Wrong And It Put The Worst Lead First

Description and use

The lead scoring system misranked companies due to missing data and an inappropriate scoring formula, causing sales teams to target unqualified prospects first.

Problem

A lead scoring system misranked companies because of missing data and an inappropriate scoring formula, causing sales teams to target unqualified prospects first.

Outcome

The revised pipeline achieved 88% coverage of key metrics and produced a lead list that correctly prioritized companies with both high payment potential and significant problems, improving sales targeting effectiveness.

Workflow

• 1. Identify target companies and required signals.
• 2. Build data collection pipeline with parallel workers, handling throttling and retries.
• 3. Sample and validate coverage for key metrics (domain age, sitemap).
• 4. Compute scores using weighted sum of ability to pay and problem severity.
• 5. Detect anomalies where one axis dominates the other.
• 6. Replace weighted sum with square root of product to enforce joint presence.
• 7. Re-rank companies and verify top list aligns with business goals.
• 8. Document findings and update pipeline for future use.

Evidence

Coverage went from 24 percent to 88 percent. My time estimate for that rerun was fifteen minutes. It took 179.

Source

Changing the core to the square root of the product fixed it. The same company dropped to 40 points and position 119.

Source