How to define your ideal customer profile for outbound

What an ICP is actually for.

An ideal customer profile is a description of the company that gets the most value from what you sell, and therefore buys fastest, churns least and refers others. Note the word company. An ICP is account-level: industry, size, stage, geography, the shape of the business. It is the answer to one question, who should we be spending our outbound time on, and just as importantly, who should we not.

That second half is where the value sits. A loose ICP is the most expensive mistake in outbound because it multiplies. Every wrong account on your list costs a researched message, a follow-up, a slot in the calendar, and a setter's judgement that gets blunted by working bad-fit leads. Get the ICP wrong and a brilliant message still books nothing. Get it right and average copy still books meetings.

Companies that aim outbound at a sharp ICP feel it across the funnel: faster sales cycles, higher win rates, better retention. The teams we work with that have a clear profile consistently see a meaningfully larger share of pipeline come from outbound, because the effort lands on accounts that can actually say yes.

ICP and buyer persona are not the same.

These two get used interchangeably and it causes real damage. The ICP is the company. The buyer persona is the person inside it. You build a list from the ICP, then you write to the persona. Confuse them and you either target individuals with no account context, or you target accounts and then send the same message to a junior analyst and a VP.

The other difference: most companies have one ICP and several personas. Inside a single ideal account you might be writing to the economic buyer, the champion who feels the pain daily, and the technical gatekeeper. Same company, three very different openers.

Build it from real accounts, not a wish list.

The single biggest mistake is defining your ICP from who you wish would buy, rather than who already does. Your closed-won deals are the most honest data you have. Start there.

Mine your best customers

Pull your ten to fifteen best accounts. Not the biggest logos, the best: the ones who closed quickly, stuck around, expanded and referred others. List them out and hunt for the attributes they share. Industry, size, business model, what triggered the purchase, who signed off. The pattern that repeats across the best accounts is your ICP draft.

Weight the analysis, then talk to a few

Aim for roughly 70% hard data and 30% conversation. The data tells you what is true across the base; a handful of customer interviews tell you why they bought and what nearly stopped them, which is the part a spreadsheet hides. Ask how they found you, what problem pushed them to act, and who else was in the room when they decided.

One useful gut-check: if 30% or more of your customers came through referrals from other customers, you are likely close to product-market fit, and your existing base is a reliable mirror of your ICP. If that number is low, lean harder on the few accounts that genuinely thrive.

The attributes that actually matter.

A usable ICP is built from a small set of attributes, each one specific enough to filter on. Vague entries (“mid-market”, “tech-forward”) are not criteria, they are vibes. Pin every line to something you can search for.

• Firmographics. Industry, employee count, revenue band, company stage, geography, business model. The hard skeleton of who is in.
• Technographics. The tools they already run. A complementary stack signals fit; a competitor's tool signals a switch conversation.
• The job to be done. The specific problem your best customers were trying to solve, in their words, not your feature list.
• Buying shape. Who owns the budget, who signs off, how long the cycle runs. This tells your setters who to actually write to.
• Disqualifiers. The attributes that reliably predict a bad fit. Often more valuable than the inclusion criteria.

Resist the urge to add twenty attributes. The strongest ICPs are tight: four or five filters that a teammate could apply to a list without you in the room. Every extra criterion you cannot operationalise is decoration.

The half of the ICP most teams skip.

Here is the distinction that turns an ICP from a slide into a pipeline engine. Your attributes split into two kinds, and most teams only ever build the first.

Filters decide who is in: industry, size, geography, the stable facts that rarely change. Signals decide when to reach out: a funding round, a new leader in the function you sell into, a hiring spree, a competitor switch, headcount growth. Filters are a standing definition. Signals are a clock.

An ICP built on filters alone gives you a static list that goes stale within a fortnight and looks, from the prospect's side, exactly like a bad scrape. An ICP layered with signals tells you which accounts on that list are receptive right now. That is the difference between “a VP of Sales at a 200-person SaaS company” and “a VP of Sales at a 200-person SaaS company who joined six weeks ago and is hiring three reps.” Same filter. Wildly different timing.

This is the core of how Occura works a list. The filters define the universe; live signals decide who gets a message this week, written by a human who can reference the actual trigger. It is also why our prospect lists are rebuilt continuously rather than scraped once and run for a quarter.

Turn the profile into a list that books meetings.

An ICP only earns its keep when it becomes a list. Take your filters into a tool like Sales Navigator or a signal provider, pull the accounts that match, then identify the right persona inside each one. That is the bridge from definition to pipeline.

From there, the work is relevance per prospect, not volume. A well-defined ICP makes the message easier to write, because you already know the problem, the buyer and the trigger. To see what a healthy funnel off a clean list looks like, the complete outreach guide walks through the numbers from a thousand prospects down to booked meetings.

And revisit the whole thing roughly every quarter. Markets move, your best-fit accounts drift, a new segment starts closing faster than the old one. An ICP that never updates slowly stops describing reality, and your list quietly degrades with it.