Lead Scoring for Sales Reps

Q: How many closed deals do I need to build a reliable model?

Start with 50 closed-won and 50 lost deals. That's enough to spot patterns and weight signals. Below 30, you're guessing. After 100, you hit diminishing returns unless you segment by deal size or industry.

Q: Should I include engagement metrics like email opens and website visits?

No. They're noise. Reps see a prospect who opened six emails but never replied and dismiss the score. Focus on signals tied to urgency and buying power: response speed, budget clarity, decision-maker level, timeline.

Q: What if my sales cycle is very long (6–12 months)?

Weight timeline and decision-maker engagement more heavily. Long-cycle deals close on persistence and stakeholder alignment, not speed. Add signals for multi-threaded contact and meeting cadence.

Q: How do I handle leads from inbound vs. outbound?

Build separate models if your close rates differ. Inbound often responds faster but may be less qualified. Outbound takes longer but often has higher ACV. Score each cohort separately, then combine if the model is stable.

Your sales reps see a lead score and ignore it. Not because they're stubborn—because the score doesn't predict whether they'll close the deal. It predicts whether marketing thinks the prospect is warm. Those are different problems. Lead scoring fails at adoption for one reason: it's built backward. Marketing teams train models on engagement metrics—opens, clicks, form fills, time on site. Then they hand the scores to sales and wonder why reps don't act on them. A prospect who opened six emails and visited your pricing page three times might never close. A quiet prospect who asked about implementation and confirmed a budget might be a done deal waiting to happen. The fix is surgical: rebuild your scoring model to predict closed won deals , not marketing funnel velocity. Then weight the signals that actually move your close rate—company size, response speed, budget clarity, decision-maker engagement. Test it against your own data. You'll find that your reps start using it within weeks. Why your current lead score is marketing theater Most lead-scoring systems are built on leading indicators that correlate with engagement, not conversion. A prospect downloads a whitepaper, visits the blog three times, and clicks a CTA button—score goes up 15 points. But engagement doesn't close deals. Budget, authority, and urgency do. The problem compounds because marketing and sales have different definitions of a "qualified" lead. Marketing sees qualification as "this person is interested enough to consume our content." Sales sees it as "this person can write a check and has a problem we solve." When you're measuring engagement, you're optimizing for the marketing definition. Your reps see leads that look hot on paper but go nowhere in practice. A prospect who opened six emails but never discussed budget or timeline isn't warm—they're a tire-kicker. A prospect who asked two questions about implementation, confirmed they have budget, and said "let's talk next week" is a deal, regardless of how many emails they opened. This gap kills adoption. Reps follow a score once or twice, get burned by a high-scoring flop, and stop trusting the system. After that, they rely on gut feel and past relationships—which are faster and feel more real. Reverse-engineer scoring from your closed-won deals Start by extracting your last 50–100 closed-won deals. For each one, pull the data you already have: Company size (employee count, ARR, headcount in their function) Response speed (days from first contact to first meaningful reply) Budget clarity (did they mention a number, a range, or leave it blank?) Decision-maker engagement (did you talk to the person who signs off, or their subordinate?) Conversation count before close (emails, calls, meetings) Implementation timeline (how many times did they mention when they wanted to go live?) Product knowledge (did they reference your product by name? Ask feature-specific questions?) Competitor mention (did they say who else they were evaluating?) Now pull 50–100 deals that fell out at different stages: proposals sent but not signed, demos booked but no follow-up, qualified opportunities that went dormant. Grab the same data. Compare the two cohorts. You'll see patterns immediately. Won deals probably responded within 48 hours. Lost deals often went silent after the second email. Won deals talked to the CFO or founder; lost deals stayed with the coordinator. Won deals mentioned a go-live date; lost deals never did. These patterns are your scoring model. Not a black box. Not a consultant's template. Your actual data. Weight the signals that predict your close rate Don't use a generic model. Build one calibrated to your business, because different sales motions have different predictors. A PLG company selling $99/month software has different close drivers than a B2B services firm selling $50K/year contracts. Start with these core signals and score them on a 0–100 scale based on your data: Response speed (0–20 points) Replied within 24 hours: 20 points Replied within 48 hours: 15 points Replied within one week: 10 points Took longer or haven't replied: 0 points Response speed is one of the strongest predictors of close rate because it correlates with urgency and priority. A prospect who replies fast has already deprioritized other options mentally. Budget clarity (0–20 points) Stated a specific number ("we've budgeted $40K") or range ("$30–50K"): 20 points Mentioned they have budget but no range: 10 points Never mentioned budget: 0 points Budget is a gate, not a gradient. Without it, you're selling into a void. Prospects who name a number have already done internal math. Decision-maker engagement (0–25 points) Direct engagement with the economic buyer (CFO, founder, VP of function): 25 points Engagement with a strong influencer (director, senior manager) who reports to the buyer: 15 points Engagement with a coordinator, analyst, or individual contributor: 5 points No identified decision-maker yet: 0 points