Lead Scoring
A methodology for ranking prospects on a numerical scale based on their fit with your ICP and the strength of their engagement or buying signals.
Lead scoring is a methodology that ranks prospects on a numerical scale based on two dimensions: fit (how well they match your Ideal Customer Profile) and engagement (their demonstrated interest through website visits, content downloads, and intent signals). Higher scores trigger sales follow-up; lower scores stay in nurture. A well-calibrated model lifts SDR conversion 2–3× and reduces time wasted on unqualified leads.
Definition
Lead scoring assigns numerical values to leads across two independent dimensions — fit and engagement — and combines them into a prioritization score. Fit scoring (sometimes called 'demographic' or 'explicit' scoring) weights attributes like job title, company size, industry, revenue, and geography. Engagement scoring (also called 'behavioral' or 'implicit' scoring) weights actions like pricing-page visits, content downloads, webinar attendance, email opens, and return visits. Modern lead scoring is often stored as a matrix (a high-fit + high-engagement lead is A1; low-fit + low-engagement is D4) rather than a single score, because the two dimensions need to be acted on differently.
Fit score vs engagement score: use both
Fit score measures how much you want the lead; engagement score measures how much they want you. High fit + high engagement is the holy grail — fast-track to an AE. High fit + low engagement is worth a nurture campaign because the account is valuable if you can wake them up. Low fit + high engagement is a support problem disguised as a sales lead; it often signals a poor ICP match worth investigating. Low fit + low engagement should be auto-filtered or unsubscribed.
Collapsing both into a single score obscures these distinctions. A lead scoring 75/100 because of huge engagement but mediocre fit is nothing like a lead scoring 75/100 because of perfect fit but mild engagement. Store fit and engagement separately, then route on the combination.
How to build a scoring model that actually works
Start with outcomes, not opinions. Pull your last 200 closed-won deals and 200 closed-lost deals. Measure which attributes and behaviors differed statistically between the two groups. Those differences become your scoring weights. A pricing-page visit that's 4× more common among closed-won leads is worth more than a webinar registration that's equally common in both groups.
Avoid the three common traps. First, don't let the model drift — re-calibrate weights every quarter as your product and market change. Second, don't confuse activity with intent — 10 blog-post views can mean a curious student, not a buyer. Third, set decay rules — a pricing-page visit from 6 months ago should not count like a visit from yesterday. Engagement without recency is noise.
Predictive lead scoring and AI models
Traditional rule-based scoring (e.g. '+10 for pricing page, +5 for whitepaper') works, but it requires manual tuning and struggles with interaction effects. Predictive lead scoring — offered by HubSpot, 6sense, MadKudu, and similar — uses machine learning to find patterns in your closed-won data and assigns weights automatically. It tends to outperform rule-based scoring once you have at least 500 closed deals to train on.
Predictive models are not magic. They need clean CRM data, statistically meaningful volume, and regular retraining. Below 500 wins/year, simple rule-based scoring calibrated to your funnel data typically outperforms ML. Above that threshold, predictive scoring starts to reveal non-obvious patterns — especially around interaction effects between firmographics and engagement behavior.
Why it matters
Sales reps have finite hours. Without a scoring model, those hours get spread across every new lead — including the ~60% that will never convert regardless of rep effort. Lead scoring concentrates that time on the top 20–30% of leads where effort actually moves the needle. Teams with calibrated models routinely report 2–3× SDR conversion improvements and 30–50% reductions in sales cycle length. It also finally aligns marketing and sales on what 'qualified' actually means.
Examples
- +10 points for viewing pricing page
- +5 points for downloading a case study
- +20 points for visiting 5+ pages in one session
How Bullseye helps
Bullseye dramatically improves the engagement half of your scoring model. Traditional engagement scoring relies on form fills and email opens — which capture only the ~3% of leads willing to identify themselves. Bullseye's person-level website intent feeds real-time behavioral signals (pricing-page views, competitor-comparison page views, return visits, dwell time) into your scoring model for the other 97%. The result: a scoring model that accurately scores the full funnel, not just the form-fillers.
Frequently asked questions
What is lead scoring?
Lead scoring is a methodology that ranks prospects numerically based on two dimensions: fit (how well they match your ICP) and engagement (their demonstrated interest through signals like website visits, content downloads, and intent data). Higher scores trigger sales follow-up; lower scores stay in marketing nurture. The goal is to focus rep time on the 20–30% of leads most likely to convert.
How do you build a lead scoring model?
Start with historical data: analyze your last 200 closed-won and 200 closed-lost deals and measure which attributes and behaviors differ statistically. Those differences become your scoring weights. Store fit and engagement as separate scores rather than one combined number, set decay rules so stale signals don't count like fresh ones, and re-calibrate weights quarterly as your product and market evolve.
What's a good lead score threshold?
There's no universal number — thresholds depend on how many leads you want to route to sales. A common starting point: the top 20–25% of scored leads become MQLs and get sales handoff; the next 30% enter accelerated nurture; the rest stay in broad nurture. Validate by measuring conversion rate by score band; adjust thresholds until MQL-to-SQL conversion consistently exceeds 30%.
What's the difference between rule-based and predictive lead scoring?
Rule-based scoring uses manually defined weights (e.g. '+10 for pricing page visit'). Predictive (AI-driven) scoring uses machine learning to find patterns in your closed-won history and weight leads automatically. Predictive tends to outperform rule-based once you have 500+ closed deals to train on; below that volume, rule-based scoring calibrated to funnel data is usually more reliable.
How does website visitor identification improve lead scoring?
Traditional engagement scoring relies on form fills and email opens — which capture only the small fraction of leads willing to identify themselves. Website visitor identification (like Bullseye) reveals behavioral engagement from the other ~97% of traffic, feeding person-level intent signals (pricing-page visits, return visits, competitor-comparison views) into your scoring model for leads who never fill a form.
Should you score at the lead level or account level?
Both, and separately. Lead-level scoring drives individual outreach and MQL handoff. Account-level scoring aggregates signals across the full buying committee — useful for ABM and multi-threaded sales. An account with five people each showing moderate engagement is often a better opportunity than a single lead with extreme engagement, and only account-level scoring surfaces that pattern.
Related terms
Buyer Intent Data
Behavioral signals — website visits, content consumption, search activity, and review-site research — that reveal which accounts are actively evaluating a purchase.
Ideal Customer Profile (ICP)
A concise description of the type of company that gets the most value from your product and returns the highest lifetime value to your business.
Marketing Qualified Lead (MQL)
A lead who meets both fit and engagement thresholds agreed between marketing and sales, making them ready for sales follow-up.
Sales Qualified Lead (SQL)
A lead that sales has formally accepted as worth pursuing after qualifying fit, need, budget, and timing — the handoff from marketing to pipeline.
Keep learning
Related Use Cases
Lead Generation
Generate leads from your website without forms by identifying anonymous visitors.
Sales Intelligence
Real-time intelligence on website visitors for proactive sales outreach.
Account-Based Marketing (ABM)
Identify individuals from target accounts visiting your site for ABM programs.
Related Guides
Put lead scoring into practice
See how Bullseye helps with lead scoring and more.