Lead scoring: the operator's guide
Lead scoring is the operating discipline of assigning numeric values to leads so sales focuses on the highest-likelihood-to-close prospects. Done well, lead scoring is the difference between sales chasing every form fill and sales talking to qualified buyers. Done poorly, it's a vanity metric that creates false confidence.
What lead scoring actually does
Two dimensions usually combine into a single score:
- Fit score (demographic). How well the lead matches your ICP. Industry, company size, role, geography.
- Intent score (behavioral). What the lead has done. Pricing page views, demo requests, content downloads, email engagement.
Mature programs maintain both scores separately and combine into a routing rule: high fit + high intent → SDR outreach now. High fit + low intent → nurture sequence. Low fit + high intent → educational content, no SDR. Low fit + low intent → minimal touch.
How to build a scoring model
- Identify your highest-value closed-won customers from the last 12-24 months.
- Identify the behavioral signals they exhibited before closing: which content, which pages, which time-to-conversion.
- Identify the demographic signals: industry, company size, role.
- Build a scoring rubric: +X points for each signal. Higher scores = higher probability.
- Validate the model against historical data — does scoring correlate with closed-won rate?
- Set thresholds: MQL at score X, SQL at score Y, immediate-SDR at score Z.
- Iterate quarterly based on what's converting.
Where lead scoring lives
- HubSpot Marketing Hub Professional+ includes lead scoring.
- Marketo has long had strong lead scoring.
- Salesforce Pardot for Salesforce-centric stacks.
- Custom-built scoring in warehouse + reverse ETL works for sophisticated teams.
- Predictive scoring (ML-based) via 6sense, Demandbase, MadKudu.
Common mistakes
- Scoring without validating against actual conversions. The model is wrong; nobody knows.
- Treating "MQL" as the conversion event. It's not — closed-won is the conversion event.
- Sales-marketing misalignment on what qualifies. The scoring threshold needs joint ownership.
- Static scoring that doesn't decay. A lead that hasn't engaged in 60 days isn't as hot as one engaged yesterday.
- Ignoring fit. High-intent SMB prospects aren't worth chasing if your ICP is enterprise.
Do I need lead scoring?
If you have a sales team and inbound lead flow, almost always yes. Below 50 leads/month it may be premature; above 200/month it's essential.
Fit or intent — which matters more?
Both, but fit usually weighs more heavily because high-fit leads close at 5-10x the rate of low-fit leads regardless of intent.
What's a good threshold?
Calibrate against your actual conversion rates. A common pattern: MQL threshold catches the top 20-30% of leads; SQL threshold catches the top 5-10%.
Operating checklist
- Define the business outcome before building.
- Audit the existing state honestly.
- Build the foundation before the advanced layer.
- Establish ownership and operating cadence.
- Measure what compounds, not what looks good.
- Refresh quarterly based on what's working.
- Document so the next operator can pick it up.