Lead Scoring Rules That Actually Work: A Practical Framework

Most sales teams don't have a lead problem. They have a scoring problem. When every lead in the pipeline looks the same on a list, reps default to working whoever came in most recently or shouting loudest — not whoever the rules flag as hot. Lead scoring turns that gut call into a number: LeadFlow flags 70+ as hot and emails at 80+ when delivery is configured. That is a priority, not a close-rate.

You don't need machine learning for this. A rule-based system with five or six well-chosen signals is easier to retune than a black-box model, because you can see exactly why a lead scored the way it did and adjust it in minutes. If you're not capturing those signals yet, our exit-intent popup playbook shows offers that match the page the visitor is leaving.

What three signal categories should you start with?

Good scoring rules pull from three buckets:

Firmographic signals describe who the lead is. LeadFlow's default set uses a corporate (non-personal) email domain as the fit proxy — not company size, industry, or job title. Those fields can sit on the lead record; they do not fire a scoring rule.

Behavioral signals describe what they've done — requested a demo, came from a webinar or referral, filled a form, arrived via Discover. These tell you intent. LeadFlow does not track page visits or email opens.

Engagement signals describe how recently they've been worked — a send, a status change, a booked call. Recency and decay are honest proxies when you don't have open tracking.

A lead with a corporate email but zero behavioral signals is worth a Pro email sequence (Day 1 → 3 → 7), not a phone call. A lead with a personal email but a booked demo still deserves a fast follow-up — they asked.

How should you assign point values?

Keep the scale simple: 0 to 100. Assign points per rule and let them add up.

LeadFlow's default set (Pro/Business) looks like this: referral +25, proposal stage +25, webinar or Discover +20, high deal value +20, paid ads or LinkedIn or corporate email +15, qualified +15, website form or contacted or recent engagement +10, 30-day decay −10. There is no pricing-page or email-open rule — LeadFlow does not track those.

Notice the negative rule at the end. Scoring systems that only add points drift upward forever and stop being useful. A decay rule for inactivity keeps the score honest — a lead that went quiet three months ago shouldn't still be flagged 70+. When a hot lead does go quiet, a five-email ask-for-the-call playbook can ask for a 30-min Product Demo — LeadFlow sequences (Pro) log Day 1 → 3 → 7 Email accepted; they do not book the call.

When should you retune the hot cutoff?

LeadFlow's UI treats 70+ as hot on pipeline cards and the scoring dashboard. High-score email alerts fire when a lead crosses 80 and delivery is configured. Those cutoffs are not a Settings slider — if you want a different watch, add a Pro automation with a score_threshold trigger and pick the number.

Resist the urge to retune rules in week one. You need at least a few dozen scored leads before you can tell whether 70 is catching the right ones or letting good leads sit at 65.

If reps are complaining that "hot" leads aren't converting, the fix is almost never a wholesale rebuild — it's usually one rule that's weighted wrong. Pull the last 20 leads at 70+ and look at what they had in common. That's your signal for what to adjust.

How often should you review scoring rules?

The single most common mistake with lead scoring is setting it up once and never touching it again. Buyer behavior shifts, your product changes, and a rule that mattered six months ago (like a webinar signup) might be irrelevant now. Put a 15-minute review on the calendar once a month: which rules are firing most often, which 70+ leads actually closed, and whether the point values still feel right.

If you're doing this in a spreadsheet, that review means manually recalculating scores. A rule-based scoring engine that runs on Pro/Business — LeadFlow's scoring works this way — just means the review is about reading a report instead of rebuilding one.

Start with five or six rules, not twenty. A scoring system with too many inputs becomes as hard to reason about as no system at all. Get the basics working, watch what actually predicts a closed deal for your business, and add complexity only when a specific gap shows up. For capture playbooks, browse the rest of the LeadFlow blog.