Stop Calling Cold Leads: How to Build a Predictive Lead Scoring Model

Your sales team is calling cold leads while hot ones go untouched. Lead scoring fixes this. Here's how to build a model that actually works.

Explicit vs. Implicit Data

Lead scoring uses two types of data:

Explicit Data (What They Tell You)

Company size (from form)
Industry (from form)
Job title (from form)
Budget (from form)
Timeline (from form)
Source: Forms, surveys, sales conversations

Implicit Data (What They Show You)

Website behavior (pages visited, time on site)
Email engagement (opens, clicks)
Content consumption (which resources downloaded)
Social media activity (LinkedIn profile views, etc.)
Source: Tracking, analytics, behavioral data

The Power of Implicit Data:

Implicit data is often more accurate. Someone might say they have a $100K budget (explicit), but their behavior (visiting pricing page 5 times) shows they're serious (implicit).

Assigning Points for Behavior (Page Views, Downloads)

Here's how to score behavioral signals:

Website Behavior Scoring

Visited homepage: +1 point
Visited pricing page: +10 points
Visited case studies: +5 points
Visited contact page: +15 points
Spent 5+ minutes on site: +5 points
Returned to site 3+ times: +10 points

Content Engagement Scoring

Downloaded ebook: +5 points
Downloaded case study: +10 points
Downloaded pricing sheet: +20 points
Watched demo video: +15 points
Attended webinar: +25 points

Email Engagement Scoring

Opened email: +1 point
Clicked email: +3 points
Clicked pricing link: +10 points
Clicked booking link: +20 points
Replied to email: +25 points

Negative Scoring: When to Deduct Points

Not all behavior is positive. Deduct points for negative signals:

Negative Behaviors

Unsubscribed from emails: -50 points
Marked email as spam: -100 points
No engagement in 90 days: -10 points
Visited careers page (probably not a buyer): -5 points
Bounced from website immediately: -2 points

Decay Over Time

Points should decay if no activity:

After 30 days of no activity: -5 points
After 60 days: -10 more points
After 90 days: -20 more points
Keeps scores current and relevant

The "MQL Threshold": When to Alert Sales

Set clear thresholds for when to involve sales:

Scoring Thresholds

0-25 points: Cold lead, nurture only
26-50 points: Warm lead, add to nurture sequence
51-75 points: Marketing Qualified Lead (MQL), notify sales
76-100 points: Sales Qualified Lead (SQL), assign to sales rep
100+ points: Hot lead, immediate sales contact

Automation Rules

Set up automation based on scores:

Score reaches 50 → Add tag "MQL", send to sales team
Score reaches 75 → Add tag "SQL", assign to sales rep, create task
Score reaches 100 → Add tag "Hot Lead", send SMS to sales, create urgent task

Iterating Your Model Based on Win Rates

Your scoring model should improve over time:

Monthly Review Process

Pull list of leads that converted
Analyze their scores at time of conversion
Identify patterns (what behaviors predicted conversion?)
Adjust point values based on actual conversion data
Test new scoring rules

What to Look For

Leads with score 50-75 that converted → Increase points for those behaviors
Leads with score 75+ that didn't convert → Decrease points or add negative signals
Common behaviors of converted leads → Increase those point values
Behaviors that never lead to conversion → Remove or decrease points

Conclusion

Lead scoring helps you prioritize. Use both explicit and implicit data, assign points for positive behaviors, deduct for negative ones, set clear MQL thresholds, and iterate based on actual conversion data. The result? Sales focuses on hot leads, and you close more deals.