How to find your first ten customers

Guide · Founder · 5 min read · last verified 2026-07-27

Your first ten customers are the buyers you win before you have case studies, brand recognition, or a repeatable channel — which means you win them one conversation at a time, using assets that do not scale: your history, your attention, and your evidence. This phase has a bad reputation because founders keep trying to skip it, reaching for automation and volume before anything is worth automating. It goes better when you treat it as what it is: a workbench phase, where every customer is built by hand and the point of the work is as much what you learn as what you earn.

Name the phase honestly

The first-ten phase has different physics from everything that follows. There is no funnel yet, only individual people. There is no conversion rate worth computing, only conversations that went somewhere or did not. Success here is not a repeatable machine; it is ten real accounts plus a written understanding of why each one said yes — the jig you will later use to shape a repeatable motion. Founders who accept this tend to make better decisions than founders who simulate scale early, because simulated scale burns the one resource this phase runs on: personal credibility. Ten hand-won customers who feel well served become your first references. Ten targets blasted with automation become closed doors.

It follows that the founder does the selling. Not because founders are natural salespeople — many are not — but because this phase is research running on a sales schedule, and the findings have to reach the person who can change the product. The reasoning is laid out in How to run founder-led research; the short version is that a delegated first-ten phase tends to produce revenue without learning, which is the more dangerous of the two failures.

Warm paths, defined widely

Warm paths tend to come first, and most founders define warm too narrowly. Warm does not mean friends; friends are usually poor first customers because they buy for the relationship. Warm means anyone for whom you arrive with borrowed trust: former colleagues who watched you work, people you have helped in communities, readers of things you have written, friends-of-customers from a previous venture, the person your first believer forwards you to. The practical move is to write the list before judging it — every person who knows your work and touches the problem — and then rank by how close they sit to the pain, not by how comfortable the message feels to send.

One caution: warm paths forgive weak targeting, and that forgiveness is a trap. A warm buyer who does not really have the problem will often buy anyway, then quietly churn, teaching you nothing except a false lesson. Warmth tends to get you the meeting; the problem still has to be real.

Cold outreach that carries evidence

When warm paths thin out — and for ten customers they usually do — cold outreach is next, and the honest version of it is research-first. You have no case studies, so the substance of the message has to come from somewhere else: the buyer's own world. That substance is gatherable — by hand in a spreadsheet at this size, or with an evidence-first platform such as Magrios once the question list outgrows you. Outreach that opens with something true and specific — a question their peers are asking, a shift in their market, an observation about their current setup — tends to be read differently from outreach that opens with your product, because it demonstrates work instead of claiming value. The full craft is covered in How to write outreach that cites your research.

What nobody can honestly promise you is a response rate. Anyone who does is selling something. What you control is relevance: whether the message could only have been written to this person. In the first-ten phase, that standard is affordable, because you need conversations in the tens, not the thousands.

What to trade when you have no proof

Early buyers accept obvious risk, and they know it. The workable posture is to price that risk honestly and pay for it with things you genuinely have:

• Access. The founder answers, directly and fast. For many early buyers, a direct line to the person who can change the product is worth more than polish.
• Attention. Onboarding done with them, not sent to them. Their edge cases get investigated, not queued behind a roadmap.
• Influence. A real say in what gets built next — offered honestly, meaning you also say no out loud when their request fights the direction.
• Price honesty. A lower early price, framed as payment for risk and feedback, with a stated path to standard terms so the arrangement has an end.

The one thing never to trade is truth. No invented case studies, no logo walls implying customers you do not have, no "companies like yours are seeing" claims you made up. Beyond the ethics, it is bad strategy: your first ten tend to talk to each other eventually, and the whole phase runs on being believed.

Ten conversations become a pattern

Treat every conversation — won, lost, or ghosted — as a data point and write it down while it is fresh: who they were, what they were doing about the problem before you, what nearly stopped the deal, what words they used for the pain. Selling and validating are the same activity at this stage, which is why this article and How to validate demand before building describe the same weeks from two angles. Somewhere in those notes, a pattern tends to surface: three or four customers who look alike, bought for the same reason, in the same words. That cluster is the exit from the workbench phase. The questions those look-alike buyers keep asking become the seed of your repeatable motion — How to choose your first buyer questions picks up exactly there — and the first structured days of pursuing a market deliberately are covered in What to do in your first week in a new market. The first ten are not a smaller version of the next hundred. They are the instructions for finding them.