Guide · B2B sales
ICP scoring criteria for B2B sales: a simple 7-point rubric
Most founders don't lose deals at the close — they lose weeks selling to companies that were never going to buy. A ten-minute ICP score fixes that. Here are the seven criteria that matter, a 0–2 rubric you can copy, and a worked example.
What is ICP scoring?
Your ideal customer profile (ICP) is a description of the company most likely to buy your offer, get value fast, and stick around. ICP scoring turns that description into a number: you rate every prospect against the same fixed criteria, total the points, and work the list from the top.
The point isn't precision — it's consistency. Without a score, prioritization defaults to "whoever replied most recently" or "whoever feels friendly." With one, a solo founder can run their pipeline with the same discipline as a sales team of ten.
The 7 ICP scoring criteria
You can invent twenty criteria; you'll actually use about seven. Each one below is scored 0 (miss), 1 (partial), or 2 (strong), for a maximum of 14 points.
1. Industry & offer fit
Is this the kind of business your offer was built for? Your best evidence is your own history: which past clients renewed, referred, or got results fastest? Score 2 if this prospect looks like them, 0 if you'd be experimenting.
2. Company size & budget realism
Big enough to afford you, small enough that you can reach the owner of the problem. A prospect who has to raid another budget line to pay you is a 0 no matter how enthusiastic they sound.
3. Pain severity
How much does the problem you solve actually cost them — in money, time, or risk? A pain they can quantify ("we lose two deals a month to slow follow-up") scores 2. A pain they merely nod along to scores 0.
4. Urgency or trigger event
Something changed: they hired, lost a client, raised money, missed a quarter, got a new mandate. Trigger events are the difference between "interesting" and "let's talk this week." No trigger, no urgency — score accordingly.
5. Access to the decision maker
Can you get the person who signs into the conversation within one or two steps? Selling through a messenger doubles your sales cycle and halves your win rate. Direct line to the owner or budget-holder = 2.
6. Deal size
Given their size and the scope they'd need, what would this engagement be worth? Score against your own floor: 2 if it's comfortably above your minimum viable deal, 0 if you'd be discounting to make it work.
7. Success & retention likelihood
If they buy, will it work? Do they have the team, the volume, or the follow-through to get value from what you deliver? A client who churns in a quarter costs more than a prospect who never buys.
The rubric
| Criterion | 0 — miss | 1 — partial | 2 — strong |
|---|---|---|---|
| Industry & offer fit | New territory for you | Adjacent to past wins | Looks like your best clients |
| Size & budget | Can't afford you | Would stretch | Budget clearly exists |
| Pain severity | Nice-to-have | Acknowledged, unquantified | Quantified cost |
| Urgency / trigger | No event | Vague "this year" intent | Recent trigger event |
| Decision-maker access | Blocked by gatekeeper | One intro away | Direct contact |
| Deal size | Below your floor | At your floor | Comfortably above |
| Success likelihood | Likely to churn | Needs hand-holding | Set up to succeed |
Reading the total (out of 14):
- 11–14 — priority. Research the account, personalize the outreach, and rehearse the call before you make it.
- 7–10 — standard. Worth a sequence, not worth an afternoon of research.
- 0–6 — disqualify politely. The deal you don't chase is profit too: it funds the hours you spend on the 11+ accounts.
Worked example
Say you run a bookkeeping service for trades businesses. Two leads come in the same morning:
Lead A — a 12-person electrical contractor whose office manager just quit (trigger, 2), owner emailed you directly (access, 2), revenue ~£1.4M (size, 2), exactly your niche (fit, 2), admits the books are three months behind and a VAT deadline is looming (pain, 2), needs your full monthly package (deal size, 2), has an operations lead to hand things to (success, 2). Score: 14. Drop everything.
Lead B — a solo consultant who "might need help someday" (trigger, 0; pain, 1), budget unclear (size, 1), not your niche but close (fit, 1), is the decision maker (access, 2), would take your smallest package (deal size, 0), organized enough to succeed (success, 2). Score: 7. Standard sequence, no bespoke effort.
Without the score, Lead B — who replied enthusiastically — probably gets your afternoon. With it, Lead A does. That swap, repeated weekly, is most of what "better pipeline discipline" means.
Using the score before the call, not just before outreach
The score also tells you which conversation to rehearse. An 11+ account deserves a practiced call: the discovery questions that surface the quantified pain, and a calm answer to the objection you know is coming. That's exactly what practicing sales calls against an AI buyer is for — you paste in the account's situation, run the call once against realistic pushback, read the scorecard, and then dial for real.
Common ICP scoring mistakes
- Too many criteria. Past ten, scoring becomes a chore and stops happening. Seven is enough signal.
- Scoring aspiration instead of evidence. "They could be a great client" is a 1, not a 2. Score what you can point to.
- Never disqualifying. If nothing ever scores below 7, your rubric is flattery, not a filter.
- Set-and-forget. Revisit the criteria quarterly against who actually bought and renewed. Your real ICP is revealed by revenue, not by brainstorming.
Related guides: build the conversation itself with a talk track, and go deeper on discovery with the Sandler pain funnel.
Scored the account? Now rehearse the call.
Paste the prospect's situation into The Sales Workout, practice against an AI buyer that pushes back like they will, and walk into the real call warm.