What Actually Predicts Lead Conversion?
An analysis of 3 years of sales data revealed surprising patterns.
Most sales teams prioritize leads based on intuition and surface-level signals. When we trained a model on actual conversion outcomes, it contradicted years of assumptions about what makes a "good" lead.
The Common Belief
Sales teams often assume that wealthier prospects make better leads. Higher property values, bigger budgets, more engagement activity - these seem like obvious predictors of success.
What the Data Showed
Analysis of 3 years of sales outcomes revealed that property value was actually a negative predictor. What mattered most: project urgency, ability to pay, and prior success in the lead's neighborhood.
The Implication
Intent signals (urgency, capacity) outperform engagement signals (email opens, page views). When scoring was rebuilt around these insights, success rates improved by 31% compared to standard CRM-based scoring.
vs. standard Salesforce lead scoring, by prioritizing intent signals over engagement metrics
Engagement Metrics vs. Intent Signals
Most lead scoring tools prioritize engagement - email opens, page views, form fills. These metrics are easy to track, but they measure activity, not intent. A tire-kicker can look engaged while a serious buyer stays quiet until they're ready to move.
Engagement Signals
What most CRMs measure
These show activity, but not commitment. A tire-kicker can score high on all of these.
Intent Signals
What actually predicts conversion
These reveal readiness to buy. A quiet lead with strong intent signals often closes faster.
Why Does This Matter?
Most CRM scoring tools default to engagement metrics because they're easy to track automatically. But engagement measures interest, not readiness. The challenge is that intent signals often require combining CRM data with external sources - property records, census data, market trends - that standard tools don't access.
When we rebuilt scoring around intent signals instead of engagement, conversion rates improved by 31% compared to standard Salesforce lead scoring.
Test your intuition
Two leads. One will convert, one won't. Can you tell which one based on property value alone? The answer might surprise you.
Which Lead Will Convert?
Test your intuition against the data
Why Intent Signals Outperform Engagement Metrics
When we trained models on 3 years of actual conversion data, the results contradicted common assumptions. Here's what the data revealed about what actually predicts a sale.
What the Data Revealed
Signals ranked by how well they predict actual conversion. Click any row to understand why.
The Key Insight
The top 3 predictors all relate to readiness to act: urgency (timeline commitment), capacity (ability to pay), and local success (reduced perceived risk). Demographic factors like income and property value - the "obvious" signals - ranked at the bottom.
Combining Multiple Data Sources
The key difference from standard CRM scoring: pulling in external data that reveals intent signals your CRM doesn't capture on its own.
Where the Data Comes From
Combining internal CRM data with external sources that reveal intent
Learns from your actual sales outcomes
Pull together CRM data with external sources for a complete picture of each lead
Apply patterns learned from past wins and losses to predict who's most likely to convert
Push prioritized rankings back to your CRM so sales sees them immediately
Why external data matters: Your CRM knows what leads tell you. External sources reveal what they don't - financial capacity, local market conditions, demographic patterns that correlate with buying behavior.
Scoring That Learns From Your Business
The difference between static rules and adaptive scoring: one approach treats every business the same, the other learns what actually works for yours.
Deep Historical Analysis
Before building anything, we dig into your historical sales data. Custom research tools help us understand what's worked in the past: which leads converted, which stalled, and why. This analysis surfaces patterns that generic tools miss because they don't know your business.
Model Tailored to Your Reality
Every business has unique conversion drivers. A roofing company's best predictors differ from a SaaS company's. We build models specifically for your situation, trained on your data, reflecting your market dynamics. No generic templates.
Continuous Feedback Loop
Scoring doesn't stop at deployment. We track which predictions were right and which missed. As new conversion data comes in, we retrain models to capture shifting patterns. Markets change, buyer behavior evolves, and scoring adapts with it.
What Any Sales Team Can Apply
Whether or not you build custom scoring, these insights can improve how you prioritize leads today.
Prioritize Intent Over Activity
A lead who says "I need this done next month" is more valuable than one who opens every email but never commits to a timeline.
Question Your Assumptions
The data showed property value was a negative predictor. What "obvious" signals might be misleading your team?
Look for Ability to Act
Urgency without capacity stalls. Capacity without urgency delays. Look for both: ready to move AND able to pay.
Leverage Social Proof
Neighborhood penetration mattered more than expected. Past wins in an area create trust that accelerates new deals.
Found this useful?
Share it with your sales team, or explore more research from our work helping organizations make better use of their data.