The rubric, straight up
Here is the whole answer, and then the reasoning. In a 100-point lead scoring model, company revenue maps to four scores: a company inside your ICP revenue band gets +25. A company one band adjacent (small enough or large enough that it could still buy, just not your center) gets +10. A company clearly out of band, one you can't serve profitably at any price, gets −10. And a company whose revenue you simply don't know gets 0. Not −5, not "pending." Zero.
| Revenue band | Score | What it means |
|---|---|---|
| Inside your ICP band | +25 | The companies your product and pricing were built for. Full firmographic credit. |
| One band adjacent | +10 | Could buy, could grow into the ICP. Real fit, weaker economics. |
| Clearly out of band | −10 | Too small to afford you or too large to serve. The negative actively pulls them below your MQL line. |
| Unknown / unenriched | 0 | Absence of data is not evidence of bad fit. Fix enrichment; don't guess. |
The 25 / 10 / −10 / 0 revenue-band rubric in a 100-point model.
Why revenue gets a quarter of the model — and no more
Revenue band is the strongest pre-intent signal you have. A lead can download every asset you publish and still never buy, because the company behind the email can't afford the product or doesn't have the problem at their size. That's why firmographic fit deserves real weight. But it's a ceiling, not a floor: 25 points means an in-ICP company still has to do something before it crosses an MQL threshold: visit pricing, book a demo, come back a third time. Give revenue 40 or 50 points and you'll route companies to sales because of what they are, not what they did. The rubric holds firmographics to a quarter of the total so behavior still decides.
Set the bands before you set the points
The numbers only work if the bands are honest, and the bands come from your closed-won data, not your ambitions. Pull your last 20 closed-won deals and note the revenue of each account. The band where deals close fast and retain well is your +25. One step outside it on either side is your +10. Everything else is the −10. If your center of gravity is companies at $1M–$10M in revenue, a $600K company that can't fund the retainer and a $200M company with a six-person procurement committee both belong in the −10 band: opposite reasons, same score. The rubric doesn't care why a company is a bad fit, only that your delivery economics say it is.
A worked 100-point model
Here's the rubric in context, in the shape I build scoring models for B2B SaaS clients. Firmographics carry 40 points (revenue band 25, industry or vertical 15), person-level fit carries 15 (role and seniority), and behavior carries 45 (engagement 25, high-intent actions like pricing visits and demo requests 20). MQL sits at 60. Run the math and the design intent falls out: an in-ICP company (+25) with a decision-maker (+15) is at 40 before anyone clicks anything. One real burst of intent puts them over. An out-of-band company (−10) mathematically cannot MQL on curiosity alone; they'd need nearly every behavior point on the board. That asymmetry is the whole point of the negative score.
| Component | Weight | Example signals |
|---|---|---|
| Revenue band | 25 | The 25 / 10 / −10 / 0 rubric above |
| Industry / vertical | 15 | Core vertical +15, adjacent +5, excluded −10 |
| Role & seniority | 15 | Economic buyer +15, influencer +8, student/consultant 0 |
| Engagement behavior | 25 | Return visits, content depth, email engagement |
| High-intent actions | 20 | Pricing page, demo request, audit signup |
A 100-point model with MQL at 60. Firmographics open the door; behavior walks through it.
The mistakes that quietly break the model
Four failure modes show up in almost every scoring model I audit. Punishing unknowns: scoring missing revenue as negative quietly buries every lead your enrichment provider whiffed on, which can be a third of your list; that's an enrichment problem wearing a scoring costume. Too many bands: a model with nine revenue tiers implies a precision you cannot validate; four scores is the most granularity that closed-won data at normal deal volumes can actually confirm. Set-and-forget: bands drift as your ACV and product move upmarket; if accounts you scored −10 keep closing and retaining, the bands are wrong, not the buyers. Re-validate quarterly. Trusting the form: self-reported revenue is fiction often enough that serious models treat an enriched figure and a form answer differently, and prefer the enriched one.
Prove the rubric against revenue, not opinion
A scoring model is a forecast, and forecasts get graded. Once a quarter, pull score-at-MQL against close rate: leads that crossed at 60 to 75 versus 75 and up, and, more telling, the close rate of everything sales worked that never hit 60. If the 75+ group doesn't close meaningfully better than the 60–75 group, your weights are decoration. And if sub-60 leads close at the same rate as scored MQLs, the model is theater and sales already knows it. This validation loop is one of the checks inside the 47-point funnel scorecard — lead qualification is where more pipeline quietly dies than in any ad account.
How gRO builds scoring models
- Bands from closed-won, not vibes. The rubric gets calibrated against your actual won-deal revenue distribution before a single point is assigned.
- Wired to routing. Scores that don't change who sales calls first are decoration — the model ships connected to CRM routing and SLAs.
- Graded quarterly. Score-to-close correlation gets checked every quarter, and the weights move when the data says they should.
FAQ
Why 25, 10, −10 and 0 specifically?
The numbers are proportions, not magic. In a 100-point model, 25 is the most weight a single firmographic attribute should carry — enough to matter, not enough to MQL a company on identity alone. The −10 exists to actively pull disqualified accounts below the threshold rather than just failing to help them. If your model totals 50 points, scale everything down: the ratios are what you're keeping.
Should unknown revenue score negative?
No. Zero. A negative score for missing data punishes leads for your enrichment provider's coverage gaps, and a third of your list can sit in that bucket. Treat unknowns as an enrichment problem to fix, and let the model score what it actually knows.
How many revenue bands do I need?
Four scores — in-ICP, adjacent, out-of-band, unknown — is the most granularity you can validate against closed-won data at normal B2B deal volumes. A nine-tier model implies precision you can't prove and adds maintenance without adding routing decisions.
Where does the revenue data come from?
A firmographic enrichment provider appending revenue (or employee count as a proxy) at form-fill or list-load, with a revenue-range dropdown on the form as fallback. Enriched data beats self-reported nearly every time — forms get answered optimistically.
What MQL threshold should I use?
Start at 60 in a 100-point model — in-ICP company plus decision-maker plus one real intent action — and then let score-to-close correlation move it. The threshold is an output of validation, not a setting you pick once.
Sources cited in this analysis
- Lead scoring frameworks — HubSpot and Salesforce scoring documentation (composite of current practice)
- Score-to-close validation methodology — gRO engagement playbook, calibrated across 15 years of B2B growth work
- The Operator's Funnel Scorecard — applygro.com/audit (lead-qualification checks)