A precise description of the type of company that gets the most value from your product and generates the most durable revenue in return — defined by traits like industry, company size, tech stack, and use case.
ICP = shared traits (industry, headcount, revenue, tech stack, use case) of your highest-value customers, where segment value = ARPU ÷ monthly churn rate (LTV)
A founder segments their Stripe customers to find the ICP. Agencies: 60 customers at £40/mo ARPU with 6% monthly churn. E-commerce brands: 25 customers at £120/mo ARPU with 2% monthly churn.
Agency LTV = £40 ÷ 0.06 = £666.67; e-commerce LTV = £120 ÷ 0.02 = £6,000 — 9× more per customer. Despite fewer than half the logos, e-commerce brands are the ICP.
An ICP describes companies, not people. It is built from firmographics (industry, headcount, revenue, geography), technographics (the tools a company runs — for a Stripe-based SaaS, 'processes payments through Stripe' is often the sharpest trait), and behavioural signals like use case. Buyer personas sit one level below: the individuals inside an account who evaluate and sign off. Target companies with the ICP; message people with personas.
The best ICP is derived from billing data, not aspiration. Rank your existing customers by lifetime value, retention, and expansion, then look for the traits the top segment shares: which industries stay longest, which plan sizes upgrade rather than downgrade, which use cases reach value fastest. Your Stripe data already contains the answer — the ICP is the pattern behind your highest-LTV, lowest-churn cohort, written down.
The most common mistake is defining the ICP by volume instead of value. The segment with the most logos is often the cheapest, churniest slice of the base, while a segment with a third of the customers but triple the ARPU and a fraction of the churn is almost always the better profile. Run the analysis on LTV per segment, not customer count. The second mistake is making the ICP so broad it excludes no one and therefore focuses nothing.
An ICP is as much a disqualification filter as a targeting tool. Every non-ICP customer you close costs you twice: once in support load and roadmap distortion, and again when they churn and drag down your retention metrics. The profile also drifts as product and pricing evolve, so re-run the segment-level LTV analysis every couple of quarters — and after any major pricing move — rather than enshrining last year's answer.
For a small team, focus is the only real leverage. Nearly every metric a founder worries about — CAC, churn, NRR, trial conversion — is a blend across customer segments, and sharpening the ICP is the cheapest way to improve all of them at once: you acquire customers who convert faster, stay longer, and expand more. Selling outside your ICP does not fail loudly at the point of sale; it fails quietly six months later as churn.
In the Ebsta × Pavilion 2025 GTM Benchmark Report (655,000 opportunities and $48B of pipeline analysed), top-performing sellers were 24% more likely than the rest to disqualify non-ICP deals early, and low performers' deals were 217% more likely to slip at late stage.
An ICP describes the ideal company — its industry, size, revenue, and tech stack. A buyer persona describes an individual person inside that company, with their role, goals, and objections. You use the ICP to decide which accounts to pursue and personas to decide how to speak to the people within them. Most SaaS businesses have one core ICP and several personas.
Start from evidence, not intuition: rank your existing customers by lifetime value and retention, take the top segment, and write down what those companies have in common — industry, headcount, tech stack, and use case. Interviewing a handful of your best customers adds the why behind the pattern. If you bill through Stripe, segment-level ARPU and churn reveal the profile directly.
Early-stage SaaS should aim for one. Multiple ICPs split marketing budget, sales messaging, and product roadmap before any single segment is won. Add a second only when a genuinely distinct segment — different buying process, different value driver — proves itself in retention and expansion data, which usually happens well after product-market fit.
Firmographics (industry, employee count, revenue range, geography), technographics (the tools they already use — for example 'bills customers through Stripe'), the trigger or use case that makes them buy, and disqualifiers: the traits that predict churn. The disqualifiers are often the most valuable part, because they tell your team which deals to walk away from.
Connect your Stripe account and see your real MRR, churn, and LTV in real time — on desktop and mobile.
No credit card required · Connect Stripe in 1 click
No credit card required. Connect Stripe in 1 click.
We're building iOS and Android apps that'll bring your metrics everywhere.
Our mobile apps are currently in active development. Follow us on social media for updates and be the first to know when they launch.