Understand your customers

How do I use the customers I already have to sharpen my ideal customer profile instead of guessing?

Once you have even 15 or 20 paying customers, stop guessing and let the data draw the profile: sort them by who onboarded fastest, uses it most, churns least, and refers others. Look for the shared traits of your top decile (industry, role, trigger, size) and write your ICP from that cluster, not from your original pitch deck. Then interview a few of them to understand why they specifically stuck. This turns ICP from a hopeful hypothesis into an evidence-based description, and it should keep updating as your best customers reveal the pattern.

Go deeper

23 resources, 22 link-checked.

▶️ Video
✓ Link checked Free Beginner

When you are new to a space, your instinct is to explain your idea and hope people nod, which teaches you nothing. This YC talk is a concrete guide to running discovery interviews the right way: extract data from the person instead of pitching, and use a small set of questions that work in any industry, including one you are still learning. It pairs well with The Mom Test as the applied version you can watch before your next call.

How to Talk to Users

On Y Combinator (Startup School) by Eric Migicovsky ~25 min

How to Talk to Users
  • The interview is to extract data, not to sell: stop talking about your idea and let them talk about their problem.
  • Skip hypothetical questions (would you use this) and ask what they have actually done to solve the problem today.
  • A handful of questions works across any industry, so you can start interviewing before you are an expert in the space.
Watch on YouTube youtube.com →
🎧 Podcast
✓ Link checked Free Intermediate

If you would rather hear the thinking than read it, this is Rahul Vohra on Lenny's Podcast talking through the same product-market-fit engine in his own voice. He is honest about deliberately ignoring most feedback and building for the narrow set of users who most love the product, which is exactly the discipline of sharpening your profile from real users. It is a repeatable system you can run yourself, explained conversationally with the messy parts left in.

Superhuman's secret to success: ignoring most customer feedback, manually onboarding every user, and positioning around one attribute | Rahul Vohra

On Lenny's Podcast by Lenny Rachitsky

  • Not all feedback is equal: he weights the people who already love the product far above everyone else when deciding what to build.
  • Manually onboarding every new user was how the team learned, in person, who the product was actually for.
  • The survey-and-segment loop is a system you run repeatedly, not a one-time exercise, so your profile keeps getting sharper.
Open lennysnewsletter.com →
🎧 Podcast
✓ Link checked Free Advanced

This compilation centers on ServiceChannel's real story of narrowing 'retail' down to a small set of exact target companies defined by size, structure, and named buyer, the same discipline the short answer asks for. It is a rare case where you hear the actual criteria a company used to cut a huge vague market down to a workable ICP.

The customer segments hiding in your own data

On Not Another CEO (ICP Compilation)

  • 'Segment until it hurts,' meaning keep narrowing past the point that feels comfortable
  • Define your ICP with concrete, checkable criteria, not adjectives
  • A narrow ICP shortens sales cycles and stops teams from chasing bad fit logos
Open notanotherceo.substack.com →
🎧 Podcast
✓ Link checked Free Advanced

This episode is the audio companion to a16z's written ICP framework, useful specifically for the moment your company has grown past its first few customers and the founding team's gut feel about the ICP no longer matches who is actually renewing. It pushes back directly on founders who assume they know their ICP without ever having checked. Good to listen to alongside the written a16z piece rather than instead of it.

Do You Really Know Your ICP? Why It Matters and How to Find Out

On a16z Podcast 35 min

  • Gut feel about your ICP expires as the company grows, check it against real account data periodically.
  • A mismatch between assumed ICP and actual best customers usually shows up first in retention, not acquisition.
  • Refining your ICP is an ongoing habit, not a one time exercise you finish at launch.
Open a16z.com →
🎧 Podcast
India Free Intermediate

Aditya Tulsian built Numberz for Indian small businesses before it was acquired by Chargebee, and this episode is a blunt, India specific conversation about the mistakes founders make defining their ICP, including whether to start with the problem or the person. SaaSBoomi is a community built by and for Indian SaaS founders, so the examples and pricing assumptions in this episode actually match what you are likely building against. A good listen if most of the other resources here feel too US centric for your market.

'Ideal Customer Profile' Truth Bombs, with Aditya Tulsian

On SaaSBoomi Podcast by Aditya Tulsian 45 min

  • Decide upfront whether you are anchoring on the problem first or the person first, both are valid but different.
  • Buyer, user, and influencer can be three different people in a B2B sale, know which one you are targeting.
  • An unclear ICP shows up later as a messy or failed acquisition conversation, not just slow sales.
Open saasboomi.org →
📄 Article
✓ Link checked Free Intermediate

Turns the vague feeling of product-market fit into a number you can move. Ask users how they would feel if they could no longer use the product, then track the share who say 'very disappointed'. Under 40 percent means keep working. A test you can run on an idea long before you scale it.

How Superhuman Built an Engine to Find Product/Market Fit

From First Round Review by Rahul Vohra ~20 min read

  • The 40 percent 'very disappointed' benchmark for product-market fit.
  • Segment to your high-expectation customers and build for them.
  • Make the fit score a metric you improve quarter by quarter.
Open review.firstround.com →
📖 Book
✓ Link checked Paid Beginner

The single best thing ever written on customer conversations. It teaches you to ask about the customer's life and past behaviour, not your idea, so you can't be lied to. If a founder reads one thing before talking to a single customer, it's this.

The Mom Test

From momtestbook.com by Rob Fitzpatrick ~130 pages

  • Talk about their life, not your idea.
  • Ask about specifics in the past, not opinions about the future.
  • 'That's so cool, I'd totally buy it' is a compliment, not data, dig for commitment and evidence.
Open momtestbook.com →
📄 Article
✓ Link checked Free Advanced

Michael King runs a16z's go to market network, and this piece is aimed squarely at the moment your first guess at an ICP needs to be tested and re-cut with real data. It gives you specific questions to ask about your existing customer base to see if your assumed ICP actually matches who renews and expands. Useful once you have some paying customers and want to check instinct against numbers.

A Framework for Defining and Refining Your ICP

From Andreessen Horowitz by Michael King 12 min read

  • Look at your best existing customers first, your ICP should describe them, not an aspiration.
  • Revisit your ICP after any major shift: new pricing, a new product line, or a wave of churn.
  • Many founders think they know their ICP but have never checked it against renewal data.
Open a16z.com →
📄 Article
✓ Link checked Freemium Intermediate

Market size is meaningless until you know exactly who you're for, and Lenny turns that into a concrete, repeatable ICP exercise from real B2B operators. It grounds your SAM in an actual customer definition instead of a fuzzy segment.

How to identify your ideal customer profile (ICP)

From Lenny's Newsletter by Lenny Rachitsky ~15 min read

  • Define the specific customer you serve before you size the market, not after.
  • A sharp ICP makes your bottom-up market math credible and your GTM focused.
  • Narrowing 'who it's for' is what makes a small beachhead winnable.
Open lennysnewsletter.com →
📄 Article
✓ Link checked Free Intermediate

This is a founder team telling on themselves: they grew for years without a defined ICP and it caused real damage to their roadmap and go-to-market. It is a useful antidote if you think ICP work is a nice-to-have exercise rather than something that changes what you build and who you call. Concrete, no fluff, written by people who actually shipped a product.

Defining our ICP is the most important thing we ever did

From PostHog

  • Product-market fit and a defined ICP are linked, you struggle to get one without the other.
  • Not having an ICP doesn't just hurt marketing, it makes the whole roadmap unfocused.
  • Define your ICP earlier than feels comfortable, waiting has a real cost.
Open newsletter.posthog.com →
✍️ Essay
✓ Link checked Freemium Advanced

Elena Verna draws the line between a broad product usage ICP and a narrower sales ICP, and shows how to build both from behavior data you already have instead of assumptions. If you have paying customers using the product differently, this essay gives you the vocabulary and method to split them into segments worth treating differently. It is written for someone with real usage data, which is exactly your situation, though the back half of the piece is gated for paid subscribers.

I'd Segment That...

From Elena Verna, Elena's Growth Scoop

  • Your self serve ICP and your sales worthy ICP are usually not the same group
  • Segment by actual product behavior, not just firmographics
  • Use segment data to decide where human sales effort actually pays off
Open elenaverna.com →
📖 Book
✓ Link checked Paid Intermediate

The best book on positioning against competitors by understanding the alternatives customers actually consider. Directly turns competitive research into a market advantage.

Obviously Awesome: How to Nail Product Positioning so Customers Get It, Buy It, Love It

From aprildunford.com by April Dunford ~200 pages

  • Your real competition is often the status quo, not another product
  • Start positioning from the competitive alternatives your buyers consider
  • Differentiated value comes from your unique capabilities vs those alternatives
  • Choose the market frame that makes your strengths obvious
Open aprildunford.com →
📄 Article
✓ Link checked Free Intermediate

This explains exactly how to build the retention and behavioral cohorts your short answer relies on: grouping customers by signup date, plan, or usage pattern to see who actually sticks. It is a practical, tool agnostic explanation of the mechanics behind sorting by who churns least, which you can apply in a spreadsheet even without Amplitude itself. Read it for the concrete cohort types, not just the concept.

Cohort Retention Analysis: Reduce Churn Using Customer Data

From Amplitude

  • Acquisition cohorts show the when of retention, behavioral cohorts show the why
  • Revenue cohorts reveal how different paying segments retain differently
  • Comparing cohorts side by side is how a churn resistant segment becomes visible
Open amplitude.com →
📄 Article
✓ Link checked Free Intermediate

This lays out how to bucket users into occasional, core and power tiers by actual product usage, which is the mechanical version of finding who uses your product most, one of the four sorts your short answer names. It also makes the case that power users aren't just active, they're your highest lifetime value and best source of referrals and feedback. Useful as a template for the usage frequency cut of your own customer list.

User segmentation: A guide to understanding your customers

From Mixpanel

  • Segment users by frequency of your core value action, not just login counts
  • Power users have outsized lifetime value and are a disproportionate referral source
  • User segmentation surfaces what your most valuable customers actually have in common
Open mixpanel.com →
📄 Article
✓ Link checked Freemium Advanced

This guide explains how to read a retention curve across different customer cohorts to spot which segment is actually sticking around long term, versus which one just looks good in the first week. It gives you the method for the churns least half of your sorting exercise with more rigor than a simple churn percentage. Worth it once you have at least a few months of usage history to plot; some depth requires a Reforge account.

Evaluate retention cohorts

From Reforge

  • A flattening retention curve for a cohort signals real product market fit for that segment
  • Compare curves across cohorts, not just a single blended average
  • The best retaining cohort is a strong candidate for your written ICP
Open reforge.com →
📄 Article
✓ Link checked Free Beginner

A step-by-step guide to running discovery when you have almost no data, aimed squarely at first-time founders. It is practical on finding people to talk to and asking in a way that surfaces truth rather than encouragement. A solid starting point before your first batch of interviews.

A Founder's Guide for Successful Early-Stage Customer Discovery

From First Round Review by Jeanette Mellinger 15 min read

  • Start discovery before you have a product.
  • Recruit the right people to talk to.
  • Ask so customers can safely tell you no.
Open review.firstround.com →
📄 Article
✓ Link checked Free Intermediate

Supan's High Expectation Customer concept adds a layer your data sort alone won't catch: among your best fit customers, find the most discerning ones, because winning them tells you the most about what your product is actually good at. She shows how she found this person concretely at Airbnb and Dropbox, which makes the abstract idea easy to copy. Read it after you've found your top decile, to sharpen who within that group matters most.

What I Learned From Developing Branding for Airbnb, Dropbox and Thumbtack

From Julie Supan, First Round Review

  • The most demanding customer in your best segment reveals your real value fastest
  • Winning your most discerning customers creates the strongest word of mouth
  • Identify this customer from real behavior, not from who you imagined at launch
Open review.firstround.com →
✍️ Essay
✓ Link checked Free Beginner

Seibel's whole point is that a market only counts once real customers are grabbing the product as fast as you can make it and paying for it, not because a report says the category is big. That is the exact test for your situation: a huge paper market with almost nobody paying is the opposite of the hair-on-fire demand he describes. Read it as a gut-check on whether you have found a real problem or just a large-sounding one.

The Real Product Market Fit

From Michael Seibel (Y Combinator) by Michael Seibel ~4 min read

  • Real product market fit shows up as customers desperate to use and pay for the product, not as a big total-addressable-market slide.
  • If usage is calm and nobody is paying, you almost certainly have not found the burning problem yet, however large the market looks.
  • A small group of customers who love you beats a large group who kind of like you.
Open michaelseibel.com →
📄 Article
✓ Link checked Freemium Intermediate

This is a clean, practical writeup of the same 40% test methodology, with more detail on how to segment survey respondents after the fact to find your strongest fit customers. It's a good companion to the original Superhuman essay if you want the process spelled out as a checklist rather than a narrative. Useful once you already have the survey results in hand and need to know what to do with them; full detail may require a Reforge account.

Measure and improve product/market fit with the 40% test

From Reforge

  • Run the very disappointed survey question, then segment the responses immediately
  • Build for the segment that's already very disappointed, not the average respondent
  • Retest periodically since your PMF segment shifts as the product changes
Open reforge.com →
🛠️ Tool
✓ Link checked Freemium Intermediate

Once you have a hunch about who your best customers are, you need to check it against behaviour, and Mixpanel is a real analytics tool with a free tier that early startups can actually use. This guide shows how to group users into cohorts by what they did and then watch which cohorts stick around, so the segment that retains best surfaces on its own. That retaining cluster is usually where your ideal customer profile is hiding.

Cohort analysis: how to read the chart, choose a platform, and turn retention into growth

From Mixpanel by Mixpanel

  • A cohort is just a group of users who share a trait or a behaviour, and cohort analysis tracks how each group behaves over time.
  • Retention curves by cohort show which type of customer actually sticks, which is stronger evidence than any single vanity metric.
  • Mixpanel's free tier makes this reachable for a small team building outside the big startup hubs, no data warehouse required.
Open mixpanel.com →
🛠️ Tool
✓ Link checked Free Intermediate

The free tool for running the how-would-you-feel-without-this survey that underpins the Superhuman method, giving you a concrete number to track once you have early users. It turns a fuzzy sense of whether people care into a repeatable measurement you can watch over time. Use it when you have enough users to survey.

Product/Market Fit Survey

From Sean Ellis and GoPractice by Sean Ellis Free tool

  • The forty percent very disappointed benchmark is a practical fit signal
  • A simple survey question turns a vague feeling into a trackable number
  • Segment the answers to learn who your product is really for
Open pmfsurvey.com →
📋 Template
✓ Link checked Free Beginner

A fill-in worksheet that pushes you past anyone who could pay into specific, written traits: company size, budget, urgency of problem, and buying process. Useful as a working document to bring to your next team meeting when you need to force a decision on who you are actually building for.

Ideal Customer Profile (ICP) Guide and Worksheet

From HubSpot

  • Write down your ideal customer's specific traits, do not leave it as a mental impression only you carry
  • Revisit and tighten your ICP as your product and market understanding evolves
  • A written ICP becomes the filter you use to say no to customers who are the wrong fit
Open offers.hubspot.com →
📋 Template
✓ Link checked Free Intermediate

This blueprint walks through analyzing your existing customer base with a structured framework to find the patterns that signal genuine fit, impact and long term value, then turning that into an ICP document your whole team can use. It is more structured than a single essay and closer to a step by step worksheet you can run this week against your own customer list. Good pick if you want a repeatable process rather than just inspiration.

How to Uncover Your ICP

From Winning by Design

  • Analyze existing customers systematically instead of relying on memory or anecdote
  • Look for patterns across situation, pain, impact and decision process, not just industry
  • Turn the pattern into a shared document your product and sales teams both use
Open winningbydesign.com →

Terms in this answer

Ideal Customer Profile Pitch Deck

People also ask

What is an ideal customer profile and how do I define mine? 24 → How specific should my target customer be when I'm just starting? 24 → What's the difference between an ICP, a buyer persona, and a target market? 24 → How do I know if I'm targeting the wrong customer? 24 → Should I focus on one customer segment or serve everyone who'll pay? 23 → How do I build customer personas that are actually useful and not fiction? 23 →

Also in How Founders Use AI

How founders actually use AI for this, over in Market Research.

How do founders validate a startup idea with AI in a weekend? 20 → How do I use AI to map competitors, pricing, and positioning? 24 →

Also in GTM

The same ground, over in Who you sell to, and why they care, our GTM track.

What is an ICP and how is it different from a persona? People use both words for the same thing. 4 → What is positioning in plain English, and why does everyone insist we have got it wrong? 3 →
eChai Partner Brands The partnership →