Lead Scoring Without the BS: A No-Nonsense Guide to Qualifying Contacts

Forget the 47-factor lead scoring model your CRM vendor sold you. Here’s what actually happens in most companies.

Someone builds a scoring model in HubSpot. Pricing page view = 10 points. Email open = 2 points. Job title contains “CEO” = 15 points. Hit 50 points = Sales Qualified Lead. Marketing celebrates. Sales gets the new dashboard. Sales ignores it because the “high-score leads” are worse than the gut-feel ones. Six months later, nobody mentions the scoring model again, and everyone goes back to the spreadsheet.

I’ve seen this pattern at probably 30 companies. It’s almost universal.

Here’s the operator version of how scoring actually works — and the embarrassingly simple alternative that beats most 100-point models.

What lead scoring is actually for

A lead score has exactly one job: change what someone does when they look at a contact.

If the score doesn’t change anyone’s behavior, the score is decoration. Most lead scoring systems are decoration. They feel rigorous (numbers! thresholds! tiers!) but nothing in the business operates differently based on the output.

The only useful question to ask of a scoring model is this: “What action does this number trigger?”

If the answer is “…the sales team prioritizes them,” ask the follow-up: “How specifically?” If there’s no specific routing rule that pulls high-score leads in front of a human, the score is producing nothing.

The 2×2 that beats most scoring models

You don’t need 47 weighted factors. You need a binary in most cases. Here’s what a real system looks like.

The two-question model

For most service businesses and B2B companies, ask:

1. Fit: Does this contact match our ideal customer profile? (Industry, size, role, budget)
2. Intent: Has this contact shown behavior suggesting they want to buy soon? (Pricing page views, demo requests, repeated engagement)

Each is a binary. Yes or no. You get four buckets.

	Low Fit	High Fit
Low Intent	Ignore	Long-term nurture
High Intent	Educate carefully	Call them today

That’s a complete scoring system. Four buckets. Four routing rules. For most businesses, this 2×2 is more accurate and more actionable than a 100-point composite score — and it takes 15 minutes to build.

When to graduate to a point-based system

Only add complexity to the 2×2 if:

1. You have enough volume that the four buckets are too coarse (typically >1,000 leads/month entering the funnel)
2. The “high fit + high intent” bucket is filling faster than sales can handle, and you need to rank within it
3. You have reliable data on which signals predict conversion (this is the hard part — see next section)

If you’re not at that scale, stay with the 2×2. The complexity overhead isn’t worth it.

The data problem nobody talks about

The textbook approach to scoring is: “weigh signals by how predictive they are of conversion.”

This requires you to know how predictive each signal is. And to know that, you need:

• Hundreds of converted leads (so you have enough data)
• Reliable attribution back to which signals each one showed
• A control group to compare against

Most businesses don’t have any of this. So they pick weights by guess (“pricing page seems important, let’s call it 15 points”), then operate on those guesses as if they were data.

The result is a scoring model that looks sophisticated but reflects nothing more than the operator’s biases. It can be worse than no scoring, because it gives the team false confidence that “the data” supports their decisions. Translation: you’ve outsourced your judgment to a math formula that’s actually just your judgment dressed up.

If you’re going to build a point-based model, you need at least 6 months of historical data and willingness to actually compute the correlations. Otherwise stick with the 2×2.

What signals actually predict conversion

If you do have the data and want to build a point-based model, here’s what tends to predict conversion (across dozens of accounts I’ve audited).

Strong signals — worth high weight

• Direct demo or call request — by far the strongest signal. Someone who books a demo is roughly 30× more likely to convert than a passive subscriber.
• Pricing page view (multiple times in a session) — strong commercial intent. Single view is weaker; pattern of returning is much stronger.
• Reply to an email — replies indicate real human engagement at a level clicks don’t.
• Multiple stakeholders from same company — B2B signal that an actual buying conversation is happening internally.

Medium signals

• Repeat website visits over a short period — they’re researching.
• Specific page views (case studies, pricing FAQ, integrations) — topical interest.
• Form fills beyond the initial entry (calculator, ROI tool) — willing to give more data.

Weak signals — most people overweight

• Email opens — increasingly noisy due to image prefetchers. Apple Mail Privacy Protection auto-opens emails. Don’t lean on this.
• Email clicks — better than opens but still subject to bot clicks. Use cautiously.
• Generic page views (homepage, about) — minimal commercial intent.
• Social follows — popularity signal, not buying signal.

Negative signals — worth tracking

• No engagement in 60+ days — should reduce score
• Unsubscribed from one list — should reduce overall engagement weight
• Hard bounces — remove from list, not just score down

A practical implementation

Here’s the version I usually build in GoHighLevel (works in any tool with custom fields and workflows).

Field setup

Two custom fields on every contact:

• fit_score (0-100, manually set or determined by form fields)
• intent_score (0-100, updated by automation based on behavior)

Workflow logic

Trigger: contact engages with site (page view, email click, form fill)
   ↓
Update intent_score:
   - Pricing page view: +20
   - Case study view: +10
   - Email click on offer link: +5
   - Demo request: +50
   - Reply to email: +30
   - 7 days no engagement: -10

Routing logic

If fit_score >= 60 AND intent_score >= 60:
   Tag: "Hot lead"
   Notify sales (Slack/SMS)
   Move to pipeline stage: "Sales priority"

If fit_score >= 60 AND intent_score 30-60:
   Tag: "Warm lead"
   Drop into accelerated nurture sequence

If fit_score >= 60 AND intent_score < 30:
   Tag: "Cold high-fit"
   Drop into long-term nurture

If fit_score < 60:
   Tag: "Low fit"
   Drop into educational content list (no sales effort)

Four tags. Four routing rules. Two scores. The scoring informs the routing, and the routing changes what humans do.

What gets measured

After 90 days running this, track:

• % of “Hot lead” tags that converted to customer (target: 30%+; below 15% means your fit definition is wrong)
• % of “Warm lead” that got upgraded to “Hot lead” within 60 days (target: 20%+)
• Sales team’s time spent on each bucket (should be concentrated on Hot, not spread)

If those numbers look right, the scoring is doing its job. If not, the weights are off and you need to retune.

Five mistakes that quietly kill scoring models

Scoring for vanity instead of action. “Our average lead score is up 12% this quarter!” means nothing if your conversion rate didn’t move.

Not decaying scores. A contact who scored 80 in January and has been silent since March should not still be an “80.” Build in time decay or behavior-based reset.

Scoring without sales feedback. The sales team is the ground truth on whether your score is right. If they say “the Hot Leads you’re sending us are mostly bad,” your model is broken — regardless of what the dashboard says.

Treating scoring as set-and-forget. Buyer behavior changes. Markets change. Your offer changes. Rescore (or at least re-validate weights) every quarter.

Overengineering before you have the data. A 47-factor model with no historical validation is not better than a 2×2. It’s worse — because it gives you confidence you haven’t earned.

What to do this week

Pick the 2×2 (Fit × Intent) model and apply it manually to your last 50 leads. See if the buckets feel right.

Make sure your CRM captures enough data to do the 2×2 automatically — at minimum: industry/role/company size (Fit) and demo request / pricing page view / email reply (Intent).

Build one routing rule: “Hot bucket” → immediate notification to sales. That single rule will usually be the highest-ROI scoring work you ever ship.

Wait 60 days. Measure conversion by bucket. Then decide whether to add complexity.

So which bucket is currently being ignored at your company — the high-fit / low-intent leads sitting in a list nobody nurtures, or the high-intent / low-fit ones getting sales attention they don’t deserve? Fix one this week. The dashboard can wait.

---

Related reading:

• Marketing Automation Fundamentals — the strategic context
• What Is Marketing Automation? A Practical Primer — basics if you’re new to automation
• Email vs SMS vs Multi-Channel — channel choices once a lead crosses a threshold
• Why Marketing Automation Fails Quietly — including the failure mode where scoring is decoration