How to Find Your First Customers Before You Build

A founder spends four months building a tool for independent consultants. The product is sharp. The onboarding is clean. Launch day arrives. They post on a few Slack groups, tweet about it once, and wait.

Nothing happens.

Not because the product is broken. Because they never figured out where consultants actually gather, what they complain about when nobody is selling to them, or how to start a conversation that wasn't a pitch. The biggest surprise on launch day wasn't that nobody bought. It was that they couldn't find the people they had built for in the first place.

I've watched this play out more times than I can count. Some of the products were genuinely good. The discipline of finding customers before building was just missing. And without that discipline, even excellent products stall.

This article is about that missing discipline. Not marketing. Not growth hacking. The simple, unsexy work of going to where your future customers already are and having conversations before you've built anything.

If you're staring at an idea right now, the question isn't "how do I market this later." The question is "do I know where to find the people who would care if this existed." If the answer is no, you have work to do this week. If the answer is yes, you probably don't need this article.

Key takeaways

• The biggest mistake founders make is asking the wrong first question. "Where do I find customers?" starts you in channels and funnels. The better question is "where are people already trying to solve this problem?" — that lands you inside the rooms where your future customers already are.
• An Early Adopter is not someone who likes new things. It is someone who has a problem they've already committed effort to solving. They reply quickly, ask follow-up questions, and have already built workarounds (spreadsheets, manual workflows, expensive alternatives).
• The channels that consistently surface early customers are Reddit, Slack/Discord communities, LinkedIn (for B2B), Facebook Groups (for specific demographics), industry newsletters, and offline meetups. Pick one or two and go deep — breadth is procrastination.
• The conversations that produce evidence are interviews, not pitches. Ask about the last time the problem came up, the workarounds they tried, and what they actually spend money on. Never ask "would you buy this?" — the answer is always yes.
• A one-week customer-discovery plan — pick a customer, find three communities, read three hours, send five outreach messages, run three conversations, summarize patterns, decide — produces a clearer picture than three months of building.

Why this matters

The most common reason a launch lands in silence is not a bad product. It is that the founder never figured out where the people they were building for actually gather, what they complain about when nobody is selling to them, or how to start a conversation that isn't a pitch. The discovery work in this article is what closes that gap. It is also what most founders skip, because it is invisible work — no commits, no deployments, no screenshots to share.

The cost of skipping it shows up at launch, and it compounds. The first failed launch teaches you to ask questions. The second failed launch teaches you to listen. By the third, you have either built the discipline in this article or you've quietly stopped trying. Customer discovery is the cheapest way to learn the lessons that would otherwise cost you three launches.

Yibud's perspective

The "find your first customer" question is the one Yibud hears most often from solo founders. It is rarely the literal question they ask. It usually arrives as "how do I market this?" or "how do I get my first ten users?" — which is the same question disguised as a tactics question. The platform's distribution score treats channel-narrowness and ICP-specificity as the first-class variables, because those are the variables that decide whether the first-customer search takes a week or a quarter.

Yibud does not replace the conversations in this article. It tells you, before you start the conversations, which channels are most likely to surface the kind of customer who matches your ICP. The conversations are still the work. The platform just helps you point them in the right direction.

If you'd like the distribution dimension applied to your specific ICP and idea, Yibud's startup validation analysis returns the highest-signal channel as a starting point in five minutes.

The Biggest Mistake Founders Make

Most founders ask the wrong question first.

"Where do I find customers?"

It's the natural question. It's also the wrong place to start. When you ask it, your brain reaches for advertising channels, social media tactics, growth funnels — all the things that come after you've identified who you're looking for. The framing assumes you already know the audience. You usually don't.

The reason this is the default question is itself worth examining — founders reach for the loud, visible half of the work (channels, tactics, funnels) and avoid the quiet half (talking to strangers, admitting the problem might not exist). If that pattern sounds familiar, the founder-psychology piece goes deeper on the founder psychology behind it.

The better question is:

"Where are people already trying to solve this problem?"

That question leads you to the rooms where people are already solving it badly — Reddit threads, Facebook groups, Slack communities, competitor reviews. When you find those rooms, you don't need to "find customers" anymore. You're already in them. The people there are solving the problem right now, badly. Some would switch if something better came along. Some would help you build it. A few might even pay you before you've written a line of code, just to make sure you ship.

This reframe isn't a tactic. It's a different starting assumption. The assumption that customers are hidden and you need to hunt them down. The reality: most future customers are sitting in public, talking about the problem you want to solve, right now. You just have to know where to look.

Who Is Your First Customer?

The first customer is not "everyone."

It's a specific person. Someone who has the problem you want to solve often enough that solving it would actually change their week. Not someone who might benefit. Not someone who could use this. Someone who currently loses time, money, or sleep to the absence of a solution.

Geoffrey Moore called these people Early Adopters in Crossing the Chasm, a 1991 book that mostly still holds up. They have a problem they're actively solving, a workaround they've cobbled together (even if painful), and they're willing to try something new and rough if it gets them closer to a real solution.

Most startup advice treats "early adopters" as a vague segment. "Tech-savvy users." "Innovators." "People who like new things." That misses the point. Early Adopters aren't defined by being early. They're defined by having a problem they've already committed effort to solving.

A consultant who spends Sunday afternoons reformatting client invoices in Excel is an Early Adopter for invoicing software. They've already built a workflow. They already feel the pain. They're already searching for a fix. You don't have to convince them the problem is real. You have to show them that your version of the fix is worth switching to.

The mistake is targeting people who might care. The friend who says "yeah, I could see using that." The colleague who nods politely. The conference attendee who takes your card. These aren't Early Adopters. They're polite audiences.

Early Adopters look different. They reply to your message quickly. They ask follow-up questions. They tell you about the spreadsheet they've built. They introduce you to someone else with the same problem. They ask when the product will exist. They're not waiting to be sold. They've been waiting for someone to build this.

Where To Look

Once you know who you're looking for, the next question is where they spend time. This isn't about scrolling through a list of popular platforms. It's about matching channels to the kind of customer you have.

Reddit is where people go to complain about problems in detail. The subreddit structure means that someone with your exact problem has probably posted about it in the last 90 days. The search bar is your friend. Type in a description of your customer's pain and see what comes up. If a subreddit has 50,000 members and active daily posts, your customer is there. If it has 200 members and one post from 2021, your customer is somewhere else.

Reddit works best where the problem is specific and recurring — personal finance, niche hobbies, professional workflows. Less ideal for problems that are private or embarrassing; some categories just don't surface there.

Indie Hackers and similar communities are good for solo founders building tools for solo founders. The audience is unusually willing to give detailed feedback because they're building themselves. The downside is that the audience is also unusually skeptical of polished marketing. Lead with substance.

Facebook Groups still work for some audiences that have migrated off Reddit. Parents. Local business owners. Hobbyist communities with a long history. Search for your customer's role plus "group" and you'll often find an active community with thousands of members. The same complaint, asked and answered, often appears in the top posts.

Slack and Discord communities are where professional tribes gather. Designers. Engineers. Marketers. Founders. Each has at least one dominant community where the same questions get asked. Joining these takes effort — you have to participate, not pitch — but the conversations are often deeper than what you'll find on Reddit.

LinkedIn is the right place for B2B audiences with a specific job title. Search for the role, follow 50 people in the role, post something useful, and watch the replies. LinkedIn rewards specificity. A post about "the bottleneck in B2B invoicing" reaches the right people. A post about "the future of work" reaches nobody.

X (Twitter) is noisy but valuable for some audiences. Indie developers. Writers. Designers. The product launch community is concentrated here. The signal is in the replies to specific accounts, not in the algorithmic feed. Find three accounts your customer respects, read the replies for a week, then contribute.

Product Hunt discussions show you what early adopters in tech are paying attention to. Even if your product isn't right for Product Hunt itself, the comments on launches reveal what matters to that audience.

Industry newsletters are underrated. Every niche has a weekly newsletter with high open rates. The writers are usually the same people who run the communities. Subscribe to five of them. Read them for a month. The problems your customers care about will start to repeat.

Offline communities still matter. Industry conferences. Local meetups. Trade shows. There's probably a regional meetup your customer attends. Show up. Don't pitch. Talk.

The point isn't to be on every platform. The point is to find the one or two places where your specific customer already gathers, and to invest real time there.

Talk Before You Build

Once you've found where your customers are, the next step is harder than it sounds: have actual conversations.

Not sales calls. Not pitches. Conversations.

Rob Fitzpatrick's The Mom Test is the best short book on how to do this well. The principle is simple: don't ask people if they would use your product. Ask them about their life, their problems, and their current workarounds. People lie about what they'd buy to be polite. They describe their actual problems much more honestly.

Steve Blank's Customer Development framework, the foundation of the Lean Startup movement, says the same thing in different language: founders are hypothesis-testing machines, and the only way to test the customer hypothesis is to talk to customers before you've decided what to build.

Good questions look like this:

• "Walk me through the last time this came up. What did you do?"
• "How often does this happen?"
• "What do you use today? What do you like about it? What do you hate?"
• "If you could wave a magic wand, what would change?"
• "Have you tried to solve this before? What happened?"

Bad questions look like this:

• "Would you use a product that does X?"
• "Do you think this is a good idea?"
• "How much would you pay for this?"
• "Would you be interested if I built it?"
• "What features would you want?"

The first set produces information. The second set produces compliments. Compliments don't help you decide what to build.

A founder I know spent two weeks asking people in his target market "would you use a tool that did X." Forty-three out of fifty said yes. He built the tool. Eight people used it. The other thirty-five were being polite.

A different founder spent two weeks asking people in the same market "walk me through the last time you had this problem." Twenty conversations. Half didn't have the problem often enough to matter. The other ten described specific workarounds. He built for those ten. Three became paying customers before launch.

Conversations matter because they reveal what people actually do, not what they think they would do. The gap between those is where most products fail.

What To Observe

Words are useful. But what people say they do and what they actually do are often different things. Watch for the signals that don't depend on what people tell you.

Repeated complaints are stronger than isolated ones. If the same complaint shows up across ten different conversations, you have a real pattern. If someone mentions a complaint you've heard three times this week, you're probably on to something.

Manual workflows are a giveaway. Spreadsheets that exist for no reason other than to track something. Sticky notes on monitors. Email folders named "TO SORT." Browser bookmarks to half-finished workarounds. People build these when the existing tools fail them. Each one is a vote for "I wish this were easier."

Spreadsheets deserve a special callout. Almost every small business operator has one they update weekly. It's a confession: existing tools didn't do what they need, so they built their own. That spreadsheet is a product roadmap written by your customer.

Workarounds are another giveaway. People copy-pasting between tools. People running the same report twice because the first one had an error. People paying a contractor $200/month to do something that should take 30 seconds. Workarounds mean the existing solutions force people to do work the software should do.

Expensive alternatives tell you the problem is real and the budget exists. If your target customer currently pays $500/month for a tool that doesn't quite solve the problem, you don't have to convince them the problem matters. They already pay to address it. You have to convince them your version is worth switching to.

Existing competitors are useful in a way most founders miss. Look at their reviews — especially the two-star and three-star reviews. Those are the closest thing to honest feedback about what the existing solution gets wrong. Each complaint in a competitor's review is a sentence someone wrote because they cared enough to be frustrated.

These signals matter more than compliments. A compliment is what someone says to be polite. A spreadsheet is what they actually do every week.

How To Know You've Found The Right People

Some conversations are polite. Some are real. The signals that separate them are worth learning.

They reply quickly. Not because they're waiting for your message. Because the problem is on their mind. People with real pain reply within hours. People who are mildly curious reply when they get around to it.

They ask follow-up questions. "When will this exist?" "How much will it cost?" "Will it integrate with X?" These questions come from people who are mentally preparing to use the thing. Polite audiences don't ask follow-up questions. They're waiting for the conversation to end.

They already spend money. On adjacent tools, workarounds, or contractors doing the work by hand. Money in the adjacent space means the problem is funded — you just need to redirect some of it.

They introduce you to others. "You should talk to my colleague." "I have a friend who has the same problem." When people introduce you to others, they're vouching for the relevance of the conversation. Polite audiences don't make introductions. They don't know who to introduce you to.

They ask when the product will exist. Not "good luck with that." Not "let me know when it launches." Actual follow-up messages checking on your progress. These are the people who will be your first customers. Treat them accordingly.

If none of these signals appear in your conversations, you're probably talking to the wrong people. That's not failure. That's information. Adjust who you're talking to and try again.

Common Mistakes

I've made most of these. You probably will too.

Talking only to friends. Your friends will say your idea is good because they love you. Their approval is not validation. It feels like validation. It's not.

Collecting compliments. "That's a cool idea" from anyone who isn't your target customer is noise. Treat compliments as data about your pitching, not about your idea.

Asking leading questions. "Would you use a tool that did X?" invites a yes. "What do you do today when this comes up?" invites an answer. The shape of the question shapes the answer.

Trying to sell immediately. Customer discovery is not sales. The moment you start pitching, the conversation stops being honest. People who would have given you useful feedback start giving you polite answers.

Ignoring negative feedback. When three different people tell you the same thing is wrong, they're right. When one person tells you something's wrong, listen carefully. The instinct to dismiss criticism is the same instinct that leads to building things nobody wants.

Searching everywhere instead of one community. Going wide feels productive. It's usually procrastination. Pick one community. Go deep. Learn it. Contribute to it. The depth beats the breadth.

Treating lurkers as customers. A subreddit with 50,000 members isn't your customer base — it's a room where some of them might be. Lurkers don't buy. Participants do.

Building before you've had five real conversations. "I'll start building and get feedback later" is the most expensive sentence in startups. Build after you've had real conversations, not before. If you can't find five people to talk to, you don't have a customer. You have an idea in search of one.

A Simple Weekly Plan

Talk is cheap. Here's a concrete plan you can execute this week.

Day 1: Identify your target customer.

Write down a specific person. Not a segment. A person. "Independent UX designers with 3-10 years of experience, freelancing, charging $100+/hour." If you can't get that specific, you don't know who you're looking for. That's day one's lesson.

Day 2: Find communities.

Identify three places this person gathers. Reddit, LinkedIn, Slack, Discord, a conference, a local meetup, an industry newsletter. If you can only find one place, the audience might be harder to reach than you thought. Three is the minimum.

Day 3: Read discussions.

Spend three hours reading. Don't post. Don't comment. Just read. Notice what people complain about. Notice what questions get asked repeatedly. Notice which posts get the most engagement. By the end of day three, you'll have a list of specific problems worth solving, in the language your customers actually use.

Day 4: Start conversations.

Reach out to five people. Not a pitch. A genuine question about their work. Use the question formats from earlier in this article. Send the messages in the morning. Check responses in the evening.

Day 5: Interview people.

Aim for three to five conversations of 20-30 minutes each. Video or phone. Take notes. Ask the follow-up questions. Listen for the signals from earlier in this article: repeated complaints, manual workflows, workarounds, expensive alternatives.

Day 6: Summarize patterns.

Write down what you heard. The themes that came up across multiple conversations. The problems that came up most often. The phrases people used to describe their pain. If you heard the same problem in three different conversations from three different people, you have a real signal.

Day 7: Decide whether to continue.

Be honest. Did you find real evidence of pain? Did you find Early Adopters with the characteristics described earlier? Or did the conversations feel like polite conversations with no real signal? Either answer is useful. If yes, keep going and start designing experiments. If no, you probably need to refine your target customer or your problem hypothesis before week two.

This plan costs nothing and takes one week. The output is a clear picture of whether your target customer exists, what they actually struggle with, and where to find them. If you skip this week, you'll spend three months building and then discover what you could have learned in seven days.

For a more structured way to think about which assumptions are riskiest in your specific idea, see the validation framework article. For a step-by-step checklist version of the validation process, Startup Validation Checklist walks through the same territory in checklist form. If your idea is a subscription product, the SaaS-specific pillar — How to Validate a SaaS Idea Before You Build It — extends the same interview discipline with a recurring-revenue and retention lens. If the product depends on AI models, the AI-specific layer — How to Validate an AI Startup Idea — adds workflow, output-quality, and model-dependency tests. For a side-by-side comparison of how the SaaS and AI stacks differ, see AI Startup vs SaaS Startup: How Validation Is Different.

Frequently Asked Questions

How many interviews are enough?

Stop when you stop hearing new information. That's usually 8-12 conversations with the right people. Five conversations with the right people is enough to learn something. Fifty conversations with the wrong people teaches you nothing. Quality matters more than count.

Should I build a landing page first?

Only if your goal is to test demand at scale. For customer discovery — finding and talking to potential customers — a landing page is a distraction. The conversations come first. If the conversations tell you the problem isn't real, you've saved yourself from building a landing page nobody converts on. If the conversations confirm the problem, then a landing page becomes useful.

Should I message strangers?

Yes. The discomfort is the point. Founders who can't bring themselves to message strangers about their idea will struggle to find customers after launch, too. Customer discovery is practice for the harder work of customer acquisition. If you can't send a polite, non-pitchy message to a stranger now, learning to do it before launch is much better than learning after.

How do I avoid sounding spammy?

Three rules help. First, lead with curiosity, not pitch. "I'd love to hear about your workflow" sounds different than "I built a tool that might interest you." Second, make the message specific to the person you're writing to. Generic templates get ignored. Specific questions get answered. Third, offer something in return. A summary of what you learned. An introduction to someone useful. Credit in a future product. Reciprocity makes outreach feel less extractive.

Can AI help find customers?

AI is useful for specific tasks in this process. It can help you draft outreach messages, summarize patterns across interview notes, and generate hypotheses about where your customer might gather. AI is not useful for the conversations themselves. You can't outsource talking to humans. The evidence of whether your idea deserves to be built still comes from real people saying real things in real conversations. Use AI to make the process faster. Don't use it to skip the process.

What if nobody in my target market responds?

Treat it as data. If you can't get three people to spend 20 minutes with you, you have a reach problem, a positioning problem, or a customer problem. Each has a different solution. Try different channels. Try different framings. Try reaching out through warm intros instead of cold outreach. If after multiple attempts nobody responds, the audience might be smaller, harder to reach, or less interested than you assumed. That's worth knowing now, not after you've built something.

What if my target customer isn't on social media?

Some customers aren't. Doctors in certain specialties. Lawyers in specific practice areas. Industrial buyers. Construction contractors. For these audiences, the channels look different. Trade publications. Industry conferences. Direct mail. Cold email. LinkedIn, which is closer to a professional database than a social network. Offline communities still exist for most professional categories. The principle is the same: find where they gather, then show up. The platform is less important than the principle.

How do I know when to stop validating and start building?

A simple test: when you have evidence — not opinions — on the four assumptions every startup depends on. Demand is real because people have agreed to pay or pre-order. Distribution is real because you can describe a repeatable channel in concrete terms. Monetization is real because the unit economics work on paper. Execution is real because you've shipped something, even small, that solves the core problem. When you have those four, build. If you're missing one, keep testing.

Where this leaves you

Most startup advice is about what to do after you've launched. This article is about what to do before.

The work is simple and uncomfortable. Find the rooms where your future customers already are. Read what they complain about. Have five real conversations this week. Listen more than you talk. Notice the signals that separate polite interest from genuine need. Decide, based on evidence, whether the problem deserves the next three months.

Good founders don't start with code. They start with conversations. The point of those conversations isn't to find users for your product. It's to discover whether the problem deserves a product at all.

If you only do one thing after reading this article, do this: send one message today to a stranger in your target market. Ask them about the last time they had the problem you're trying to solve. Listen to the answer. Notice whether the answer is real or polite. Use that answer to decide what to do next week.

If you'd like a structured second opinion on which parts of your idea are worth testing first before you commit three months of work, Yibud's startup validation analysis can help. The goal isn't to score your idea. The goal is to surface the assumptions most likely to derail you, so the conversations you have this week are the ones that matter most.