Most bad ICPs don't come from a lack of effort — they come from a handful of the same shortcuts, repeated across almost every early-stage team. 1. Starting with a person instead of a problem Picturing the customer before identifying the pain almost always produces a persona wearing an ICP's name tag. The fix is starting from severity, frequency, and cost of inaction instead — covered in our core ICP guide. 2. Writing it once and never touching it again An ICP frozen at launch

AI-generated ICP tools exist now — plug in a URL, get back an audience description in seconds. The question worth asking isn't whether they work, it's what they're actually doing when they "work." What they're genuinely good at Turning a blank page into a plausible-sounding first draft fast Summarizing your own website copy back into structured, firmographic-sounding language Giving a team something concrete to react to and disagree with, instead of starting from nothing What

The core framework doesn't change between a SaaS product and a consulting practice — but the inputs that fill it in look different enough that copying a product-company template rarely fits a service business cleanly. Trigger events look different Product businesses often key off budget cycles, a new tool in the stack, or headcount growth. Service businesses more often key off a capacity gap, a bad experience with a previous vendor, or an approaching compliance deadline nobod

Once the first ICP is written down, a second question shows up fast: do we need one for every product, every segment, every persona that shows interest? Usually not — but the honest answer is more nuanced than a flat "just one." The case for one A single ICP forces discipline. Every message, every ad, every piece of content can be built for one specific buyer instead of watered down to work for everyone a little. Most early-stage teams that struggle with generic messaging are

An ICP that lives in a slide deck doesn't buy media. The fastest real test of whether an ICP is actually useful is simple: could you hand it to whoever runs your ad account and have them know exactly what to build? What a targeting brief needs that a generic ICP doesn't Firmographic and behavioral traits translated into actual interest, lookalike, or custom-audience seed criteria — not just written as a description The trigger event translated into timing and structure — a la

Most ICP templates online are built for teams that already have a CRM full of data. This one's built for the more common situation — you have a hypothesis, a handful of conversations, and nothing else yet. Fill in what you can; let testing fill in the rest. Company profile Industry / vertical Company size — employees or revenue band Geography Tools or tech stack they already use Trigger event — what has to happen for them to be actively looking now (funding round, new regulat

Most ICP work stops at describing the dream customer. The teams that waste the least time also write down who to walk away from — on paper, before the objection comes up in a live call. What a negative ICP actually is A negative ICP is an explicit list of traits that disqualify a lead, even if it superficially resembles your ICP. It's not "anyone who says no" — it's a specific, repeatable pattern of who churns, complains most, or never should have been sold to in the first pl

An ICP isn't a document you write once. Markets shift, products mature, and the customers who were once your best fit stop looking like your future ones. Here's what usually tips a team off that it's time to revisit it. 1. Your best customers stop matching your original profile If the accounts renewing, expanding, and referring you no longer look like what's written in your ICP doc, the doc is describing your past customers, not your best ones. 2. Sales cycle length is creepi

Structured interviews are how you tell a real painkiller apart from an idea that only sounds good in the room. Why interviews beat surveys here Surveys capture opinions. Interviews, done right, surface behavior — what someone has actually tried, paid for, or built to solve the problem. Behavior is the signal; opinion is noise. Who to talk to, and how many 10 to 15 structured conversations is usually enough to see a pattern emerge. Fewer than that and you're reacting to one lo

A written ICP is only useful if it changes what your team does with an actual list of leads. That's what a scoring model is for — turning "does this company fit?" into a number sales can act on without a debate every time. Why scoring beats a gut check Without a score, every lead gets an ad hoc judgment call, which means two reps treat the same lead completely differently. A scoring model forces the criteria that matter into the open and makes prioritization consistent. A sim

"ICP," "buyer persona," and "target market" get used interchangeably in half the marketing content out there — and that's exactly why so many teams end up chasing the wrong list. They're three different lenses, zoomed to three different levels. Target market — the widest lens Target market is every company or person who could plausibly use what you sell. It's broad on purpose — the total pool, not a targeting strategy. "Small businesses that need accounting software" is a tar

An ICP that lives in a doc nobody opens doesn't move pipeline. The fastest place to put it to work is the thing most teams already do badly: cold outreach. Generic outreach fails because it skips the ICP "Hi, I saw you're in [industry], thought this might be relevant" is what a template looks like when nobody consulted the ICP. It maps to nothing specific about this account. Three things a good ICP gives your outreach A trigger event to reference — the reason this account is

Ask ten founders how they built their ideal customer profile, and nine will describe a person: age range, job title, company size, maybe a stock photo in their head. That's a buyer persona wearing an ICP's name tag. The founders who get to product-market fit faster tend to start somewhere else entirely — with the problem, not the person — and treat the whole thing as a hypothesis to test rather than a document to finish. Start from the problem, not the person When you have no

Most solar companies write their ICP backwards. They picture the customer first — homeowner, mid-40s, environmentally minded, owns a single-family house — before they've asked what problem actually gets someone to sign a contract for a roof full of panels. With zero customer data, that mental picture is a guess wearing a profile's clothes. There's a steadier place to start, and it doesn't begin with a person at all. Start from the problem, not the person When you have no cust