How to Define Your ICP: A Practical Guide with Real Examples

Your ICP is the short, testable description of the companies and people most likely to buy, stay, and refer, written specifically enough that you could hand it to a stranger and they would build the same target list you would.

Most teams think they have an ICP. What they actually have is a vague sense of "mid-market SaaS companies" or "growing agencies" that falls apart the moment someone asks who to email this week. A real ICP is not a mission statement. It is a filter. If it does not change which prospects make the list and which get dropped, it is not doing its job. We run outbound every day, and the single biggest lever on reply rates is not the copy, it is who the copy goes to. Below is how we define an ICP, refine it, and use it to decide who gets contacted.

Start from customers, not aspirations

The mistake we see most often is defining an ICP from who you want to sell to rather than who actually buys. Aspiration is fine for a pitch deck. It is poison for a target list.

Pull your last several closed deals, especially the ones that closed quickly and stuck around. Look for what they had in common before they ever spoke to you. If you have too few customers to see a pattern, use your best-fit pipeline and your fastest sales cycles instead. The goal is to reverse-engineer the shape of a good buyer from evidence you already have, not to guess.

Two customers who both "love the product" are not an ICP. Two customers who both had a specific problem, at a specific size, with a specific trigger, and closed in under a month, are the beginning of one.

The two halves: firmographic and behavioral

A usable ICP has two layers. Firmographic criteria describe who the company is. Behavioral criteria describe what they are doing right now. You need both. Firmographics alone give you a big static list. Behavior tells you which names on that list are worth a message today.

Firmographic criteria

These are the durable attributes of the account:

• Industry or vertical — be specific. "SaaS" is a category, not a target. "Vertical SaaS for the trades" is a target.
• Company size — headcount and revenue bands. A 15-person company and a 900-person company are different buyers even in the same industry.
• Geography — where they operate, and where you can actually sell, support, and get paid.
• Business model — do they sell B2B, run a marketplace, take usage-based revenue? This shapes their pain more than their logo does.
• Tech or tooling — what they already run often signals both budget and readiness.

Here is a compact example of firmographic criteria for a hypothetical outbound-tooling company:

Behavioral criteria

Firmographics tell you the account could be a fit. Behavior tells you the timing is right. This is where fit becomes intent, and the two are not the same thing, which we cover in more depth in our guide on intent vs fit.

Behavioral signals worth tracking include hiring for a relevant role, launching a new product line, recent funding, leadership changes, visible use of a competitor, or engagement with your content. Each one is a reason a good-fit account might be ready now rather than in six months. Layering these signals on top of firmographic fit is the core idea behind lead scoring, and it is what turns a static list into a ranked queue.

The practical rule: firmographics decide who is eligible, behavior decides who is urgent.

Narrowing versus broadening

An ICP is a dial, not a switch. The right width depends on where you are.

Narrow when your reply rates are low, your sales cycle is inconsistent, or your messaging feels generic. A tighter ICP lets you write a message that names a real, specific pain, and specificity is what earns replies. It is almost always better to send 100 sharply targeted messages than 1,000 vague ones. Early-stage teams especially should start narrower than feels comfortable, because a narrow ICP produces a narrow, believable pitch.

Broaden when you are consistently winning inside a segment and running out of room, or when you notice adjacent accounts converting that fall just outside your current definition. Broadening should be evidence-driven: you expand because the data showed you a new pocket of fit, not because you got impatient.

The failure mode on both ends is real. Too narrow and you cap your pipeline before you have a business. Too broad and every message reads like it was written for no one. If you cannot describe your ICP in one or two sentences that would exclude most companies, it is too broad.

Using your ICP to filter outreach

An ICP that lives in a slide is useless. It only pays off when it changes what happens in your outbound. In practice, we use it as a gate at three points.

First, list building. Every prospect either matches the firmographic criteria or does not make the list. No exceptions for "they seem interesting." This is exactly the work behind knowing your prospect before a single message goes out, and it is the difference between a targeted campaign and a spray.

Second, prioritization. Among the accounts that pass the firmographic gate, behavioral signals decide the order. An eligible account showing a live trigger jumps the queue. An eligible account with no signal waits. This is the same logic our AI SDR applies when it decides who to contact first across email, LinkedIn, WhatsApp, and voice.

Third, message selection. A tight ICP lets you write to a specific situation instead of a generic benefit. When you know the segment, the size, and the trigger, the message almost writes itself, and personalization stops being a cosmetic first-line trick and starts being genuinely relevant.

At Leaderra this is not a manual chore. The whole point of running AI agents on outbound is that ICP criteria and behavioral signals become the rules the agents actually enforce, so every contact that goes out has already passed the filter. Flows start at $500/month, and the ICP is the input that makes those flows worth running.

Refining from real data

Your first ICP is a hypothesis. The market grades it. Once outreach is live, you get a feedback loop most teams ignore.

Watch who replies positively, who books, and who closes, then compare them against your ICP. When accounts outside your definition keep converting, that is a signal to broaden or add a segment. When accounts inside your definition consistently ghost or churn, that is a signal to tighten or drop a criterion. Negative signals matter as much as positive ones. The companies that reply "not relevant" are telling you where your firmographics are wrong.

Set a cadence to revisit it, monthly or quarterly, not never. Markets move, your product evolves, and an ICP that was right a year ago quietly goes stale. The teams that win outbound treat the ICP as a living document that every campaign either confirms or corrects. If you want a structured way to score fit and feed those results back in, our lead scoring guide walks through the mechanics.

If you would rather see this run end to end, watch the live demo or book a meeting and we will walk through how ICP criteria drive real outreach.

FAQ

What is the difference between an ICP and a buyer persona?

An ICP describes the ideal company or account to sell to, using attributes like industry, size, and behavior. A buyer persona describes the individual person inside that company you need to reach, including their role, goals, and objections. You need both, and the ICP comes first because it decides which companies are worth building personas for.

How specific should my ICP be?

Specific enough that it excludes most companies in the market. If your ICP would let almost any B2B company through, it is too broad to guide targeting or messaging. A good test is whether a stranger could use your written ICP to build the same prospect list you would build yourself.

Should I have more than one ICP?

Early on, one tight ICP is almost always better than several loose ones, because it keeps your messaging sharp and your learning fast. As you scale and prove repeatable wins in a segment, you can add a second ICP for an adjacent market. Adding ICPs before you have proven the first one usually spreads your effort too thin.

How do firmographic and behavioral criteria work together?

Firmographic criteria decide whether an account is eligible to be contacted at all, based on durable attributes like industry and size. Behavioral criteria decide the timing and priority, based on signals like hiring, funding, or product launches. Fit makes an account worth targeting, and behavior makes it worth targeting now.

How often should I update my ICP?

Treat it as a living document and revisit it on a regular cadence, such as monthly or quarterly. Update it whenever your outreach data shows accounts outside the ICP converting well or accounts inside it consistently failing to engage. The market is the real grader of your ICP, so let live results drive the refinements.