Not from a channel. From you, personally, one at a time — and that is the correct way to do it.
The first ten customers come from direct, personal outreach by the founder: people you already know, people one introduction away, and specific individuals you contact one at a time with a message written for them.
Not ads. Not SEO. Not a launch.
Because channels are for scaling something that already works, and at ten customers nothing works yet. You do not know the message, the objection, the segment or the price.
And because doing it manually is how you learn all four. Every conversation teaches you something that no dashboard would have told you.
A founder wants their first ten customers. What they actually do:
Week 1 — list 60 people matching the ICP: 12 they know, 18 one introduction away, 30 found via LinkedIn and trade directories.
Week 2 — 12 warm conversations. 4 say yes to a pilot. Two objections come up repeatedly: 'we already have a process' and 'who else uses this?'
Week 3 — rewrite the outreach to address both up front. Ask the 4 pilots for introductions. Send 30 personalised messages — a specific line about that firm, not a template.
Week 4 — 9 replies, 5 calls, 3 more customers. Total: 7.
What they now have that ads would never have given them: the two objections that actually block deals, the phrasing that gets replies, and the knowledge that referrals convert at roughly triple the rate of cold outreach.
That is the real product of the first ten customers.
First-time founders often launch on a big platform and expect the first customers to arrive. A launch to an audience that does not know you produces a spike and no retention — and teaches you nothing about why people did or did not stay.
The second mistake: mass, templated outreach. Two hundred generic emails get a 1% reply rate and no learning. Thirty specific ones get 20% and a conversation.
The third: not asking every single customer for an introduction. It is the cheapest channel that exists and founders forget it constantly.
Names, not segments. Split into people you know, one-hop introductions, and cold.
They reply, they are honest, and their objections teach you what to say to everyone else.
One specific line about them proves it is not a template. That single line is most of the reply rate.
The two or three that keep recurring are your real messaging problem. Address them before they are raised.
At the moment they are happiest — right after they get value, not at renewal.
Automation locks in a message you have not validated. Do it by hand until the message stops changing.
From your first customer to roughly your fiftieth, depending on price point.
Very low-priced consumer products cannot be sold one at a time economically. Even then, the first hundred users should come from places you personally participate in, not from paid acquisition.
Apply this to your own startup in My Full Journey (free account).