Ideal Customer Profile

What an ICP is

An ICP is a decision document, not a marketing slide. It names the accounts that your product genuinely fits, the accounts that are a stretch but still viable, and the accounts you should not spend a single SDR hour on. The value is in the explicitness. A vague ICP (mid-market B2B companies) is functionally no ICP at all; a useful ICP defines the employee range, revenue range, specific industries included and excluded, required technologies, required stage, and the observable behaviors that separate in-market from out-of-market.

The structure we use has three tiers. Tier 1 ICP is the narrow core: accounts where the fit is so strong that outbound effort compounds reliably. Tier 2 is the plausible adjacent: accounts where the fit holds but the sales cycle is longer or less predictable. Tier 3 is the opportunistic surface: accounts where we will not pursue outbound but will respond to inbound. Effort allocation follows the tier, not the other way around.

What an ICP is NOT

• It is not a persona. Personas describe individuals; the ICP describes accounts. A B2B buy requires both layers, but conflating them flattens the thinking.
• It is not a marketing exercise. Sales, CS, and product all have stakes in what the ICP says. An ICP built by marketing in a vacuum fails the moment it meets the pipeline.
• It is not static. The ICP that worked at USD 1M ARR is usually wrong at USD 10M ARR, and almost always wrong by USD 50M. Revisit every 6-12 months with closed-won and churn data in hand.
• It is not a wishlist. It is a constraint. Accounts that do not fit get removed, not demoted.

How it operates inside the Growleads playbook

The ICP is the foundation of our demand intelligence build. Every downstream system (signal scoring, routing, message selection, channel choice) keys off the ICP definition. We do not start with signals. We start with the question: who are we actually built to win, and how do we know? Only once the ICP is honest do we turn on the signal capture layer.

The most common ICP failure we see is that the stated ICP and the closed-won ICP do not match. A client will say their ICP is USD 10M-USD 50M ARR fintech; the closed-won data will show 70% of their last 40 deals came from USD 5M-USD 20M ARR companies, half of which were fintech-adjacent insurtech or proptech. We call this the ICP drift. Fixing it usually means redrawing the tier-1 box smaller than the client had defined it, then expanding tier-2 to capture the adjacent category that has been quietly converting.

A concrete example

A Growleads client had defined their ICP as 500-5,000 employee B2B SaaS companies in the US. Closed-won analysis showed the real ICP was 800-2,500 employee companies, 60% vertical SaaS (not horizontal), with a specific set of technographic markers (Segment + Snowflake + a scheduling tool in the stack). Redrawing the tier-1 box against that reality reduced outbound universe by 72%. Closed deals per quarter went up 34% the following quarter, because the SDR team stopped burning effort on accounts the product never really fit.