How to Set Up Lead Scoring
Prioritize your best leads with a data-driven scoring model
Not all leads are created equal. Lead scoring helps you prioritize follow-up by ranking leads based on their likelihood to buy. This guide shows you how to build a lead scoring model that separates hot prospects from tire-kickers.
Quick answer
To set up lead scoring, decide your fit and engagement criteria, assign point values weighted by historical conversion data, set MQL and SQL thresholds with sales, implement the rules in your CRM or MAP, and layer in website behavior from visitor identification. Review scoring quarterly against actual closed-won data to keep the model predictive.
Define Your Scoring Criteria
Lead scoring typically includes two dimensions: fit (demographic/firmographic match to ICP) and engagement (behavioral signals showing interest). Define criteria for both.
- Fit factors: company size, industry, job title, location
- Engagement factors: page views, content downloads, email opens
- Weight factors by importance to conversion
- Include negative scoring for bad fit signals
Assign Point Values
Give each criterion a point value based on its correlation with conversion. High-intent actions (pricing page, demo request) should score higher than low-intent (blog view).
- Pricing page view: +20 points
- Downloaded guide: +10 points
- Job title match: +15 points
- Wrong industry: -25 points
Set Score Thresholds
Define what score triggers each action. Typically: MQL threshold for marketing nurture, SQL threshold for sales handoff.
- MQL: 50+ points (marketing qualified)
- SQL: 80+ points (sales qualified)
- Start conservative and adjust based on feedback
- Review with sales to align on thresholds
Implement in Your CRM/MAP
Configure scoring rules in your marketing automation or CRM platform. Most tools (HubSpot, Marketo, Salesforce) have built-in lead scoring capabilities.
- Set up automation to update scores in real-time
- Trigger alerts when leads hit thresholds
- Log scoring changes for transparency
- Test thoroughly before going live
Add Website Visitor Behavior
Enhance scoring with first-party intent data from website visits. Identified visitors who view high-intent pages multiple times are your hottest leads.
- Score based on specific pages viewed
- Weight repeat visits higher than single visits
- Use time on site as engagement signal
- Integrate visitor ID data into scoring model
Monitor and Refine
Lead scoring is not set-and-forget. Continuously analyze which scores correlate with actual conversions and adjust accordingly.
- Track conversion rate by score range
- Survey sales on lead quality feedback
- Adjust weights quarterly based on data
- Remove criteria that don't predict conversion
Common mistakes to avoid
- ✕Making scoring too complex at the start
- ✕Not including negative scoring for bad fits
- ✕Setting thresholds without sales input
- ✕Never updating the model after initial setup
- ✕Ignoring behavioral signals and only scoring demographics
Pro tips
- ★Start simple with 5-10 criteria, then expand
- ★Website visitor behavior is the strongest intent signal
- ★Recency matters—recent engagement should score higher
- ★Build separate models for different products/segments if needed
Tools you'll need
HubSpot
Built-in lead scoring with automation
Salesforce
Einstein lead scoring with AI
Marketo
Advanced scoring for enterprise
Frequently asked questions
What is lead scoring?
Lead scoring is a methodology to rank leads based on their likelihood to buy. It assigns point values to lead attributes and behaviors, helping sales prioritize the most promising opportunities.
What factors should I include in lead scoring?
Include both fit factors (company size, industry, job title) and engagement factors (page views, content downloads, email engagement). Weight high-intent behaviors like pricing page visits heavily.
How do I know if my lead scoring is working?
Track conversion rates by score range. High-scoring leads should convert at significantly higher rates than low-scoring leads. If not, adjust your scoring criteria and weights.
Should lead scoring be manual or automated?
Automated. Manual scoring doesn't scale past a few dozen leads a week and creates inconsistency between reps. Use your MAP or CRM (HubSpot, Marketo, Salesforce) to apply scoring rules the moment a lead is created or updated, and to re-score whenever new behavior fires. Reps should still validate high scores before outreach, but the initial ranking should happen automatically so nothing slips through the cracks.
How often should I update my lead scoring model?
Review it every quarter against conversion data. Pull all MQLs from the last 90 days and bucket them by score range—if your 80+ leads aren't converting materially better than your 50–79 leads, the weights are off. Retire criteria that no longer predict conversion, add new ones from emerging behaviors, and re-validate. A scoring model that doesn't move in a year is probably wrong.
Continue learning
Deep Dive Guides
Related Use Cases
Key Terms
Ready to start doing this?
Bullseye makes it easy to implement what you just learned.
More how-to guides
How to Identify Anonymous Website Visitors
The complete guide to anonymous website visitor tracking — turn the 97% who never fill a form into named, contactable leads.
How to Generate B2B Leads
Proven strategies to fill your pipeline with qualified business leads
How to Run an ABM Campaign
Execute account-based marketing campaigns that drive pipeline
How to Improve Your Sales Pipeline
Strategies to increase pipeline velocity, conversion, and predictability