Lead scoring is one of those things that every B2B company knows they should do, most companies attempt at some point, and almost nobody does well.
The typical lead scoring implementation goes something like this: marketing sets up a point system in their CRM. Opening an email gets 5 points. Visiting the pricing page gets 20 points. Downloading a whitepaper gets 15 points. When a lead hits 100 points, they're "sales-ready" and get routed to a rep.
Three months later, sales is complaining that the "hot leads" are garbage, marketing is pointing at the scores defensively, and nobody trusts the system. The scores get ignored, and the company goes back to gut-feel qualification.
Sound familiar? Here's why it breaks, and how to build a model that actually works.
Why Most Lead Scoring Fails
Problem 1: Activity ≠ Intent
The fundamental flaw in most scoring models is the assumption that activity equals intent. A prospect who opens every email and downloads every piece of content might be a researcher, a student, a competitor, or simply someone who likes reading. Activity without context is noise.
Problem 2: No Negative Scoring
Points go up, but they never come down. A lead who was highly active six months ago but has gone completely silent still shows a high score. This creates a pool of "qualified" leads that are actually dead, wasting sales time and destroying trust in the system.
Problem 3: Fit Is Ignored
A one-person consultancy that obsessively reads your blog and attends every webinar will outscore a VP at a Fortune 500 company who visited your pricing page once. If your model doesn't account for fit, company size, industry, role, budget, it will consistently prioritise the wrong leads.
Problem 4: No Calibration
Most companies set up lead scoring once and never revisit it. But your market changes, your content strategy evolves, your sales process shifts. A model built 18 months ago is scoring based on assumptions that may no longer be true.
The Three-Dimensional Model
Effective lead scoring operates across three dimensions: Fit, Intent, and Timing. Each dimension gets its own score, and a lead must meet thresholds across all three to be considered sales-ready.
Dimension 1: Fit Score (0-100)
Fit measures how well a lead matches your Ideal Customer Profile. This is largely demographic and firmographic:
Company attributes:
- Revenue range (e.g., $5M-$50M = 25 points, outside range = 0)
- Employee count (e.g., 50-500 = 20 points)
- Industry (target industries = 20 points, adjacent = 10, irrelevant = 0)
- Geography (serviceable markets = 15 points)
Person attributes:
- Title/seniority (decision-maker = 20 points, influencer = 10, individual contributor = 0)
- Department (relevant department = 15 points)
Fit scores are relatively static. A company either matches your ICP or it doesn't. This dimension acts as a filter, no amount of engagement should make a poor-fit lead "sales-ready."
Threshold: Leads with a fit score below 50 should never be routed to sales, regardless of their activity.
Dimension 2: Intent Score (0-100)
Intent measures buying signals, actions that suggest a lead is actively evaluating solutions like yours. This is where most companies put all their scoring eggs, but the key is weighting actions by their actual correlation with purchase:
High-intent signals (20-30 points each):
- Visited pricing page
- Requested a demo or trial
- Visited comparison/vs pages
- Searched branded terms
Medium-intent signals (10-15 points each):
- Attended a product-focused webinar
- Downloaded a bottom-of-funnel asset (ROI calculator, implementation guide)
- Visited case studies
- Returned to site multiple times in a short period
Low-intent signals (3-5 points each):
- Opened marketing emails
- Downloaded top-of-funnel content
- Followed on social media
- Visited blog posts
Critical: Decay. Intent scores should decay over time. A pricing page visit from yesterday is a strong signal. The same visit from three months ago is meaningless. We typically apply a 30-day half-life, signals lose 50% of their value every 30 days.
Dimension 3: Timing Score (0-100)
Timing measures whether the lead is in an active buying window. This is the most underused dimension, but often the most predictive:
Strong timing signals (25-40 points each):
- Recently received funding
- Posted job listings for roles related to your solution
- Underwent leadership change in relevant department
- Contract renewal date approaching (if known)
- Mentioned relevant pain points on social media or in a conversation
Moderate timing signals (10-20 points each):
- Company growing rapidly (headcount increase >20% in 6 months)
- Recently adopted complementary technology
- Competitor of existing customer
Timing data often comes from external sources, funding databases, job boards, technographic tools, news monitoring. Integrating these into your scoring model requires enrichment tools, but the payoff is significant.
Putting It Together
A lead is sales-ready when they meet minimum thresholds across all three dimensions:
Dimension | Threshold | What It Means
Fit | ≥ 50 | They match our ICP
Intent | ≥ 40 | They're showing buying behaviour
Timing | ≥ 30 | They're likely in an active buying window
A lead with Fit: 80, Intent: 60, Timing: 45 is genuinely hot, they match your profile, they're actively engaged, and the timing is right.
A lead with Fit: 30, Intent: 90, Timing: 50 should not go to sales despite high engagement, they don't match your ICP.
A lead with Fit: 80, Intent: 15, Timing: 60 goes into a nurture sequence, great fit, great timing, but they haven't shown enough interest yet.
Calibration: The Ongoing Work
Your scoring model isn't finished when you launch it. It's finished after 3-4 calibration cycles. Here's how:
- After 30 days: Review every lead that was routed to sales. Did the rep agree it was qualified? If not, why?
- After 60 days: Look at conversion rates by score band. Are high-scoring leads actually converting at higher rates? If not, your weights are wrong.
- After 90 days: Analyse closed-won deals. Work backwards, what did these leads look like before they became opportunities? Use this to validate and adjust your scoring criteria.
- Quarterly: Full model review. Drop signals that don't correlate with conversion. Add new ones based on what you've learned.
The Bottom Line
Lead scoring should be your sales team's best friend, a system that consistently surfaces the right leads at the right time. If yours isn't doing that, the model is probably one-dimensional, static, and un-calibrated.
Build across fit, intent, and timing. Decay old signals. Set hard thresholds. And above all, calibrate relentlessly. A good scoring model isn't built, it's evolved.