Skip to content

An Ideal Customer Profile is only useful if your team can disqualify against it. Otherwise it's a description of the market dressed up as a strategy.

What separates a defensible ICP from a poster on the wall

Most ICPs that get written never disqualify anyone. They describe a company that exists in theory and welcome every lead that walks through the door anyway. That isn't an ICP. It's permission to keep doing what you were doing.

A defensible ICP does three things at once. It names the firmographic and technographic profile of the companies most likely to buy and stay. It names the trigger events that move them into market. And, most importantly, it gives your team the language to say no to the accounts that look right and are actually wrong.

The data behind a Critical Deal ICP

  • Pipeline analytics. Your last 18 months of opportunities, scored by win-rate, ACV, sales cycle length, expansion potential and churn risk.
  • Firmographic overlay. Sector, size band, geography, growth stage, ownership, regulatory regime.
  • Technographic overlay. The stack they already run, the platforms they already pay for, the integrations that matter.
  • Trigger signals. Funding rounds, leadership changes, M&A activity, hiring patterns, public disclosures.
  • Primary buyer interviews. Where the firmographic data does not explain a pattern, we go talk to the buyers.

What you get

  • A named Ideal Customer Profile with the criteria that put a company in or out.
  • A tiered model (Tier 1 / Tier 2 / Tier 3) so demand-gen, sales and CS can prioritise consistently.
  • A disqualification rulebook, the explicit conditions under which a lead is bounced.
  • A trigger model, the buying-readiness signals that move a target account up the priority list.
  • An activation pack for RevOps to operationalise scoring inside your CRM.
How we run it

Five steps to an ICP your team will use.

  1. Hypothesis ICP.

    We start from your current account list and your current beliefs. Both get tested.

  2. Pipeline analytics.

    Last 18 months of opportunities, scored by win-rate, ACV, cycle length, expansion and churn.

  3. Firmographic and technographic enrichment.

    Sector, size, geo, stack and platform fit. Patterns surface fast at this layer.

  4. Primary buyer interviews.

    Where the data does not explain a pattern, we go and ask the buyers why.

  5. Operationalise.

    Disqualification rulebook, tiered targeting, trigger model, CRM scoring spec. The team can act on Monday.

If you cannot disqualify with it, it isn't an ICP.
How we tell a real ICP from a description of the market.
Frequently asked

ICP research, questions.

What is ICP research?

ICP research is the structured discovery of which companies you should be selling to, and which ones you should be disqualifying, built from primary evidence rather than from internal opinion. The output is a profile sharp enough to act on: who's in, who's out, and why.

How is ICP research different from buyer persona research?

ICP describes the company. Persona describes the human inside it. Both are needed. ICP gets you to the right account; persona gets you to the right person inside that account. We typically run them together when the company is at the segment-fit stage.

What signals go into a defensible ICP?

Firmographics (sector, size, geography), technographics (the stack they already run), behavioural signals (the triggers that put them in market), economic signals (budget, growth, funding), and the win-rate and expansion data from your own pipeline.

How small should an ICP be?

Small enough to disqualify roughly 70% of your inbound. If your ICP qualifies almost every lead that comes in, it isn't an ICP, it's a description of the market.

How long does ICP research take?

Three to five weeks. Faster if you have rich CRM data and a defined hypothesis; slower if we need to run primary buyer research alongside the firmographic work.

Who uses the output?

Demand-gen, sales, customer success, RevOps and the CFO. Demand-gen for targeting. Sales for disqualification. CS for prioritisation. RevOps for scoring. CFO for sales-and-marketing efficiency arguments.

How often should ICP be re-run?

Light refresh every 12 months. Full re-base every 18–24 months, or when growth stage changes (Series A → B, B → C, mid-market → enterprise).

Book a call → Free 30 minutes. No obligation. We reply within one working day.
Get in touch

Tell us the decision you're trying to get right.

We reply within one working day.