How to Apply ICP Criteria for Lead Scoring
Learn how to apply ICP criteria for lead scoring with a practical points-based framework — firmographic weights, a worked example, and setup steps.
Conteúdo
Most lead scoring models fail for a boring reason: they score behavior (opened an email, visited pricing) without ever checking whether the lead was worth scoring in the first place. A perfect fit who never clicks anything gets buried under a terrible fit who downloaded three ebooks.
The fix is to score fit and behavior separately, starting with fit. That means turning your ideal customer profile into a points-based system your CRM can apply automatically — which is exactly what "applying ICP criteria for lead scoring" means in practice.
This guide walks through how to do that: which criteria to score, how to weight them, a worked example with numbers, and where the underlying data actually comes from.
What Does It Mean to Apply ICP Criteria for Lead Scoring?
Applying ICP criteria to lead scoring means converting your ideal customer profile's firmographic traits — industry, company size, revenue, geography, tech stack — into a point value per lead, so reps can rank and prioritize leads by fit instead of by gut feeling or arrival order.
Which ICP Criteria Should You Score?
Not every ICP trait deserves equal weight. Start with the criteria that actually correlate with your closed-won deals, not every field your CRM happens to have.
- Industry match — does the lead's industry sit inside your defined ICP verticals?
- Company size — headcount within your target band (e.g. 50–500 employees)
- Revenue range — annual revenue inside the band you can serve profitably
- Geography — a region you can sell, support, and stay compliant in
- Tech stack — tools that signal compatibility or buying readiness (e.g. already runs HubSpot or Salesforce)
- Growth signals — active hiring, funding, or expansion that suggests budget exists
- Negative criteria — traits that predict churn or a dead deal, scored as deductions rather than additions
Most of these are firmographic data points rather than anything a rep can eyeball from a LinkedIn profile — which is why manual ICP scoring tends to fall apart at volume.
How to Build a Points-Based ICP Scoring Model, Step by Step
Step 1 — Rank your criteria by correlation, not intuition. Pull your last 12 months of closed-won and closed-lost deals. For each ICP trait, check how often it shows up on the won side versus the lost side. Traits that show up on both aren't differentiators — drop them from the model.
Step 2 — Assign a point range per criterion. Give your two or three strongest predictors a wider range (0–20 points) and weaker signals a narrower one (0–5 points). Keep the total achievable score to a round number like 100 — it makes thresholds easier to communicate.
Step 3 — Set your tiers. Define what a score means operationally: which range gets auto-routed to an SDR, which gets nurtured, and which gets disqualified. Three tiers is usually enough — more than four tends to confuse reps rather than help them.
Step 4 — Add negative criteria as deductions. If a trait reliably predicts churn or a stalled deal — under 10 employees, no outbound motion, a competitor already embedded — subtract points instead of leaving it out. A lead that scores 70 on positive traits but hits a hard negative criterion should not out-rank one that scores 55 cleanly.
Step 5 — Automate the scoring, don't run it manually. A scoring model that lives in a spreadsheet gets updated once and forgotten. Push the logic into your CRM's scoring engine (HubSpot's lead scoring properties, for example) so every new or updated lead is re-scored the moment its data changes.
Step 6 — Re-score on fresh data, not stale imports. A scoring model is only as good as the fields it reads. If headcount, revenue, or industry codes were last updated at import time, the score is scoring last year's company. Real-time enrichment — live web research at query time rather than a lookup against a database compiled months ago — keeps the inputs current, which keeps the score honest.
ICP Lead Scoring Example
Here's a simplified points model for a B2B SaaS company selling to mid-market European teams (100-point scale):
| Criterion | Points available | Example scoring |
|---|---|---|
| Industry match | 0–20 | 20 if core vertical, 10 if adjacent, 0 otherwise |
| Company size (50–500 employees) | 0–20 | 20 inside band, 10 within 50% of band, 0 outside |
| Geography (EU + UK) | 0–15 | 15 inside, 5 elsewhere but GDPR-eligible, 0 otherwise |
| Tech stack signal (runs HubSpot or Salesforce) | 0–15 | 15 if present, 0 if absent |
| Growth signal (active SDR hiring) | 0–15 | 15 if hiring, 0 if not |
| Revenue band fit | 0–15 | 15 inside target band, 5 borderline, 0 outside |
| Negative criterion (no outbound motion) | −20 | Deducted if confirmed |
A lead scoring 70+ routes straight to an SDR. 40–69 goes into a nurture sequence. Below 40 gets disqualified without a call. The exact weights should come from your own won/lost analysis, not this table — this is a starting structure, not a universal formula.
Where to Get the Data to Score Leads Against Your ICP
A scoring model is only useful if the fields it depends on are actually populated and current. This is where most teams get stuck: the CRM has an "Industry" field, but it's empty for a large share of leads and stale for the rest.
Enrich-CRM fills that gap with 250+ company-level data points and 50+ contact-level data points per record, pulled through live web research rather than a static, periodically refreshed database — so the headcount, revenue band, and tech stack you're scoring against reflect the company as it is now. Intent signals can layer on top of ICP fit to flag which fit accounts are showing buying behavior this week, and job change detection keeps scores accurate when a champion at an ICP-fit account moves roles.
To operationalize the model itself, the scoring rules can run inside HubSpot directly, be built as filters in Clay, or be triggered through Zapier, Make, or n8n — and for teams with their own stack, the same data is available through a REST API or a simple CSV workflow.
One more thing worth building into the model, especially for European teams: where the underlying data is processed. Enrich-CRM is GDPR-native, with servers in the EU (Paris) — a compliance criterion that belongs in your vendor evaluation as much as in your customer's.
FAQ
What is ICP-based lead scoring?
ICP-based lead scoring is a method of ranking leads by how closely their firmographic profile — industry, size, revenue, geography, tech stack — matches your ideal customer profile, assigning point values to each matching trait so leads can be prioritized by fit rather than by arrival order or engagement alone.
How is ICP scoring different from behavioral lead scoring?
ICP scoring measures fit (is this the right kind of company), while behavioral scoring measures intent (is this lead engaging right now). The two should be combined, not treated as substitutes: a high-fit lead with no engagement may just need a different outreach angle, while a high-engagement lead with poor fit rarely converts into a good customer.
How many points should each ICP criterion be worth?
There's no universal weighting — it depends on which traits correlate with your own closed-won deals. Run the analysis in Step 1 above before assigning point ranges; copying someone else's weights skips the step that makes the model accurate.
Can I apply ICP scoring without a data enrichment tool?
Yes, but only at small volume and only briefly. Manual research works for a handful of leads a week; beyond that, missing or outdated firmographic fields quietly break the model, since a scoring rule can't evaluate a field that's empty.
How often should ICP lead scores be recalculated?
Ideally every time a lead's underlying data changes — new funding round, new headcount, new tech stack — rather than on a fixed schedule. That's realistic only with automated, real-time enrichment feeding the CRM fields the scoring model reads from.
Ready to score leads against your real ICP instead of guesswork? Create a free Enrich-CRM account — 100 credits per month, no credit card required, paid plans start at €29/month.