How to Set Up Lead Scoring

Define Your Scoring Criteria

Start by identifying the attributes and behaviors that predict a lead will become a customer.

1. Pull a list of your last 50 closed-won deals. Look for patterns in company size, industry, job title, and technology stack. These become your firmographic and demographic scoring criteria.
2. Map the buying journey in your CRM. Identify which actions (page visits, content downloads, email replies, demo requests) happened before a deal was created. These become your behavioral scoring criteria.
3. Separate your model into two dimensions: fit score (how well the lead matches your ICP) and engagement score (how actively they interact with your brand). A lead with high fit but low engagement needs nurturing. A lead with low fit but high engagement is probably not worth pursuing.
4. Interview your top 3 sales reps. Ask them what signals make them prioritize one lead over another. Their instincts often reveal scoring criteria that data alone misses.

Score Type	Example Criteria	Points
Firmographic	Company size 50-500 employees	+15
Firmographic	Target industry match	+20
Demographic	VP or Director title	+15
Demographic	Relevant department (Sales, Marketing, RevOps)	+10
Behavioral	Visited pricing page	+20
Behavioral	Downloaded gated content	+10
Behavioral	Attended webinar	+15
Behavioral	Requested demo	+30
Negative	Personal email domain	-20
Negative	Student or intern title	-30

Build the Scoring Model

Translate your criteria into a working model inside your marketing automation platform.

1. Assign point values to each criterion based on correlation with closed-won deals. Start simple. You can refine the weights after you have data.
2. Set a threshold for MQL (Marketing Qualified Lead). A common starting point is 50 points for fit and 30 points for engagement, but calibrate this to your pipeline.
3. Add negative scoring rules. Leads with competitor email domains, unsubscribes, or job titles outside your buying committee should have points deducted automatically.
4. Create decay rules for engagement scores. If a lead has not interacted with your content in 30 days, reduce their engagement score by 20%. Stale engagement is misleading.
5. Document every rule in a shared spreadsheet so sales and marketing can review and agree on the logic.

Align Sales and Marketing on Thresholds

A scoring model is useless if sales ignores the scores or marketing sets the bar too low.

1. Define SLA response times by score. Leads above 80 points should be contacted within 2 hours. Leads between 50-80 should be contacted within 24 hours.
2. Run a pilot with 3-5 sales reps for 30 days. Have them rate each MQL they receive as “good,” “okay,” or “bad.” Track these ratings against scores to validate your model.
3. Hold a monthly calibration meeting between sales and marketing. Review the leads that converted versus those that did not. Adjust scoring weights based on actual outcomes.
4. Set a target MQL acceptance rate of 70% or higher. If sales rejects more than 30% of MQLs, your scoring model needs tightening.

Monitor and Optimize

Lead scoring is not a set-it-and-forget-it system. Plan for ongoing maintenance.

1. Review scoring accuracy quarterly. Pull conversion rates by score range and look for scoring ranges where leads convert at unexpectedly high or low rates.
2. Add new scoring criteria as your product and market evolve. If you launch a new feature, add related page visits as behavioral signals.
3. Consider moving to predictive lead scoring once you have 12+ months of data. Machine learning models can surface patterns that manual scoring misses, but they require clean historical data to work.
4. Track the time from MQL to opportunity creation. If scored leads are not converting faster than unscored leads, something in the model is broken.