Every agency has an ideal customer profile somewhere.

I bet yours lives in a Google doc that hasn't been opened since the day it was written. It describes a fluffy, overly broad version of a potential client ("growth-stage SaaS companies that value strategic partnership"). That Gdoc is wallpaper in your GTM war room when it ought to be an active tool that gives you a YES or NO and sometimes a maybe.

A real ICP is a meaningful description of firmographics (vertical, size, etc), likely buying committee members (who makes & influences the purchase decision), desired attributes (tech stack, funding status, etc) & most importantly A SCORING RUBRIC. The scoring rubric is basically how you & your sales team are deciding to swipe right (or not).

It's a tool the sales team uses to rate every prospect on the same scale, with the same criteria, in under five minutes. The output is a tier (A, B, C, D, or F) and a number that tells you what to do next. The rubric makes the ICP operational - and that changes it from wallpaper to a GTM machine.

Why Most ICP Fail at Being Useful

Three failure patterns are nearly universal in the agencies I work with:

  1. The ICP is qualitative: "Growth-stage B2B SaaS companies with marketing teams that value strategic partnership over execution." Every word is true. None of it helps a sales rep decide whether the prospect in Apollo really fit the criteria. The stick part here is that qualitative ICPs “feels” sophisticated but are mostly vibes. Rarely does a purely qualitative ICP produce operational lift.
  2. The ICP has too many criteria: We will often see a list of 15 attributes the ideal customer should have. But real prospects might match only 6, sometimes 10, and once every millenia a prospect will match all 15. With so many criteria, every prospect "kind of fits" and every prospect "kind of doesn't." When everybody sorta fits and sorta doesn't, it means that the ICP isn't functioning as an attention guidance system - pointing your team at the best opportunities.
  3. The ICP has no scoring system: Even when the ICP criteria are good, there's no way to combine them into a decision. Is a prospect that matches on industry but not on size better or worse than a prospect that matches on size but not on industry? Without weights and thresholds, there's no answer and the team defaults to gut feel. Gut feel is fine when the team has 50 prospects to evaluate per quarter. It collapses at 500.

This rubric fixes all three by giving you five criteria, explicit weights, and tier thresholds that turn the score into a recommended action.

The rubric: five criteria, weighted, on a 0-100 scale

The rubric scores every prospect on five criteria, each worth up to 20 points. Total possible score: 100.

The five criteria are:

  1. Industry fit (0-20): does the prospect operate in your target vertical?
  2. Company size (0-20): is the prospect in your revenue and team-size sweet spot?
  3. Growth trajectory (0-20): is the prospect growing, flat, or declining?
  4. Buying readiness (0-20): does the prospect show signals of being in market right now?
  5. Delivery fit (0-20): can you actually help them with your current offer and capability?

Five criteria is the right number because each one is independently meaningful and the combination produces a reliable signal. More criteria add complexity without adding signal. Fewer criteria miss patterns that predict whether you'll win the deal and whether you'll deliver well if you do.

The scoring guidance for each criterion

The point of the rubric is that two different sales reps scoring the same prospect should arrive at similar numbers. That requires explicit scoring guidance per criterion. Vague guidance produces vague scores.

1. Industry fit (0-20)

  • 20: Exact target vertical, ideal sub-segment within it. You have at least two case studies that match.
  • 15: Target vertical, adjacent sub-segment. You have at least one case study that's similar.
  • 10: Adjacent vertical with transferable patterns. Your work applies but the language and reference points are different.
  • 5: Outside your verticals but the shape of the problem is familiar.
  • 0: Wrong vertical entirely, no transferable patterns. The work you do doesn't translate.

2. Company size (0-20)

  • 20: In your revenue sweet spot, optimal team size for your delivery model. You've delivered for companies this size before and the engagement structure works.
  • 15: Adjacent to your sweet spot. A small adjustment to your engagement model lets you fit.
  • 10: Smaller or larger than ideal but workable with effort.
  • 5: Significantly outside your ideal size. The engagement will require structural changes to fit.
  • 0: Too small to afford your work or too large to make you the right buyer.

3. Growth trajectory (0-20)

  • 20: Clearly scaling, well-funded, accelerating. Public signals (funding, hiring, expansion) confirm the growth.
  • 15: Growing steadily, healthy fundamentals, no distress signals.
  • 10: Flat or modest growth. Stable but not expanding rapidly.
  • 5: Declining or restructuring. Possible recovery but requires meaningful change.
  • 0: In active distress, layoffs, bankruptcy risk. Engagement is high-risk.

4. Buying readiness (0-20)

  • 20: Active mandate, new leader in the last 12 months, recent budget event (funding round, fiscal year start, M&A activity). They are buying now.
  • 15: Clear intent signals, actively exploring options, multiple touchpoints with your category in the last 90 days.
  • 10: Problem-aware but no urgency. They know they need help eventually.
  • 5: Not currently in market. Long-term watch.
  • 0: No signal of intent, no budget visibility, no urgency. Cold prospect.

5. Delivery fit (0-20)

  • 20: Their problem matches your strongest case study and your core offer. You can deliver excellent work and produce a reference-quality outcome.
  • 15: Their problem maps to your offer with minor adjustment. You'll need to flex but the work is well within capability.
  • 10: Their problem requires you to stretch your delivery. You can do the work but it won't be your best.
  • 5: Their problem is outside your core competence. You'd be learning while you deliver.
  • 0: You cannot help them succeed with your current capability. The honest answer is to refer them elsewhere.

Tier thresholds and what each tier means

The score becomes useful when it converts to a tier and the tier converts to an action.

  • A-tier (90-100): Pursue aggressively. Founder-level outreach if you sell into founder accounts, senior-rep outreach otherwise. These are your highest-priority deals.
  • B-tier (75-89): Pursue with standard sales motion. Prioritize in the pipeline review. Allocate normal effort.
  • C-tier (60-74): Qualify further before investing time. Light-touch nurture. Move to A or B only if signals strengthen.
  • D-tier (40-59): Deprioritize. Monitor for changes in size, growth, or buying readiness. Do not invest active selling time.
  • F-tier (0-39): Do not pursue. Document why for future reference. The rubric is calibrated when F-tier prospects rarely show up in inbound, because the rubric tells you what to filter out at the top of the funnel.

The thresholds matter because they convert the score into a decision rule the entire team can apply without escalation.

Example: 3 prospects scored

Three real-shape prospects (anonymized) scored through the rubric.

Prospect 1: A $30M Shopify Plus DTC brand

You are a paid social agency that specializes in DTC brands $5M-$50M.

  • Industry fit (20/20): Exact target vertical, you have four case studies in this sub-segment.
  • Company size (20/20): Right in the sweet spot.
  • Growth trajectory (15/20): Growing steadily, recently profitable, no distress.
  • Buying readiness (15/20): New head of growth, three months into the role, actively evaluating partners.
  • Delivery fit (20/20): Their problem is the exact problem your strongest case study solved.

Total: 90. A-tier. Founder-level outreach.

Prospect 2: A $150M B2B SaaS company

You are the same paid social agency.

  • Industry fit (5/20): Wrong vertical (B2B SaaS not DTC). Some transferable patterns but most of your work doesn't apply.
  • Company size (10/20): Larger than your typical client but workable.
  • Growth trajectory (15/20): Growing steadily.
  • Buying readiness (10/20): Problem-aware, no obvious urgency.
  • Delivery fit (5/20): You'd be learning while delivering. Risk of delivering mediocre work.

Total: 45. D-tier. Deprioritize.

The score tells you the truth: this is a tempting prospect because of size, but the industry and delivery fit are wrong and you'd be selling work you can't excellently deliver.

Prospect 3: A $4M Shopify brand recovering from a difficult year

Same agency.

  • Industry fit (20/20): Exact target vertical.
  • Company size (15/20): Edge of your sweet spot, smaller than typical but workable.
  • Growth trajectory (5/20): Declining year, recovering. Possible recovery but distressed.
  • Buying readiness (15/20): Founder-led, clear intent, exploring options.
  • Delivery fit (15/20): You can help, but the engagement will be harder than typical because the company has financial pressure.

Total: 70. C-tier. Qualify further before investing time.

The score reflects the reality: the industry and intent are right, but the financial pressure on the company is a real risk that needs to be evaluated before you commit selling effort.

How to calibrate the rubric to your business

The rubric template above will be wrong for your specific agency on the first pass. Calibration is the work that makes it accurate.

Run the rubric against your last 30 closed-won deals. If your average won-deal score is below 75, your thresholds are too high or your criteria are too strict. If your average won-deal score is above 90, your thresholds are too low and the rubric isn't differentiating between good deals and great ones.

Run the rubric against your last 30 closed-lost deals. If lost deals are scoring as high as won deals, the rubric is not predictive and you need to revise the criteria.

The pattern you're looking for: won deals cluster in A and B tiers, lost deals cluster in C and D tiers. When you see that pattern, the rubric is calibrated and you can use it for decisions about real pipeline.

Calibration takes an afternoon. The output is a scoring system you can hand to a new sales hire on day one and know they'll apply it the same way you do. That's the difference between an ICP that lives on a slide and an ICP that operates.

What to do with this

Take the rubric. Replace the criteria with your actual targets. Score five prospects today. See what tier they land in. Compare to your gut feel for each prospect.

If the rubric tells you something different than your gut, the rubric is usually right. Gut is overweighting the things you noticed most recently. The rubric is weighting all the things consistently.

For the full ICP system, including the scoring framework the rubric supports and the AI scoring prompt for running this at scale, the rubric is one piece of the larger sales operating system. For the entire system, see SalesOS.