Your First 100 Users: How to Win Them by Hand

We got our first 100 users one conversation at a time. Not one campaign, not one viral thread, not one clever growth loop — a hundred separate, slightly awkward, deeply human exchanges. I want to tell you that there was a trick. There wasn't. Getting your first 100 users is unscalable, manual, and a little humbling, and after doing it I'm convinced that's not a bug in the process. It's the whole point.

The fantasy is that you ship something good and the world arrives. The reality is that nobody knows you exist, and the few who do haven't decided to care yet. Your first users don't show up. You go and get them, by hand, the way you'd recruit people for anything that matters.

Why your first users have to be won by hand

There's a piece of writing every founder eventually collides with: Paul Graham's Do Things That Don't Scale. The line that rearranges your brain is this one: "Startups take off because the founders make them take off." Not because the product is so obviously great it sells itself. Because two or three people decided, every single day, to drag it into the world.

Graham's blunt about the most common version of this work: "The most common unscalable thing founders have to do at the start is to recruit users manually. Nearly all startups have to." That last sentence is the part people skip. You are not the exception. The team you admire that "just took off" almost certainly spent a brutal, unglamorous stretch hand-recruiting strangers before the takeoff you saw.

So the first reframe is this: hand-recruiting your first 100 users isn't a sign you've failed at marketing. It's the job. Acquisition at this stage is a manual labor, and the sooner you stop waiting for it to feel automated, the sooner you have customers.

Go where the users already are

The biggest unforced error I see is founders trying to summon an audience to a place that doesn't have one yet — their own homepage, their own Slack, their own brand-new subreddit. Empty rooms stay empty. Your first 100 users are already congregating somewhere, having the exact conversation your product answers. Your job is to show up in that room as a person, not a banner.

Lenny Rachitsky studied how the biggest consumer apps got their first 1,000 users and found that nearly all of them got there through one of just a handful of strategies — and almost none of them scalable at the start. DoorDash didn't run ads. Tony Xu and his team printed flyers charging $6 for delivery and papered Stanford's campus with them, routing orders through a bare-bones site. That's not a growth hack. That's three people deciding to physically walk where their users walked.

Concretely, going where the users are means: finding the three or four communities where your buyer already complains about the problem, becoming a genuinely useful regular there for weeks before you ever pitch, and earning the right to say "I built a thing for exactly this." Slack and Discord groups, niche subreddits, the replies under the right person's posts, the comment sections of the blogs your buyer reads at midnight. You don't broadcast. You join.

DMs and warm intros beat any funnel

For our first stretch, the highest-converting channel was the most embarrassingly manual one: I wrote individual messages to individual people. Not a blast. A real note that proved I'd read their work, named the specific problem I thought they had, and asked one small question. Most got a reply. Many became users. A few became the people who told the next ten.

Warm intros are even better, because they arrive with borrowed trust. The math is simple and unforgiving early on: a cold message converts in the low single digits; a warm intro from someone the person already trusts converts shockingly well. So spend your social capital. Ask the five people who believe in what you're doing for one introduction each. That's potentially five conversations that start at "I trust the person who sent you" instead of "who are you."

This is slow. You will do the arithmetic and realize that at this rate, reaching 100 users by hand will take weeks. Do it anyway. The conversations are not overhead — they are the product research you'd otherwise pay a fortune to fake.

Onboard the first ones yourself, in person

Here's the part founders rush past on their way to "scale": the first users should be onboarded by you, personally, ideally while watching them. Concierge onboarding feels indulgent. It's the opposite. It's the fastest learning loop you will ever have.

The team behind Superhuman built one of the most admired onboarding experiences in software, and they got there by doing it entirely by hand first. Their take is unambiguous: "every founder should spend time manually onboarding their earliest customers." They ran live, white-glove sessions inspired by Apple's Genius Bar and five-star hotel concierges — dozens of specialists walking thousands of users through the product, one at a time — and only encoded those lessons into a self-serve flow once they understood exactly where people got stuck.

When you sit with someone as they try your product, you learn things no analytics dashboard will surface for months: the word on your button that confuses them, the step where their face falls, the workaround they reach for that tells you what they actually wanted. Airbnb learned this in the most literal way possible. Graham notes that the founders spent the early days going door to door in New York, recruiting users and helping them fix their listings — and discovered, among other things, that hosts were taking terrible photos. So the founders started shooting the photos themselves. You don't find that in a funnel report. You find it on someone's couch.

The unscalable things compound

The Collison brothers at Stripe became the patron saints of this approach. When someone agreed to try Stripe, they didn't email a signup link — Graham describes the "Collison installation," where they'd say "Right then, give me your laptop" and set the person up on the spot. They removed every inch of distance between interest and use, by hand, one merchant at a time.

None of this scales, and that's exactly why it works. Doing things that don't scale buys you something no amount of automation can: a small group of people who feel personally seen, who you understand at a depth your competitors don't, and who will forgive the rough edges because a human showed up for them. That's the seed crystal. Reputation early on is a function of how each individual user feels, not how many there are — and a hundred users who feel that way will carry you further than ten thousand who signed up and forgot.

The takeaway

Your first 100 users will not be acquired. They'll be recruited, one humbling conversation at a time. Go to the rooms where they already gather. Write the message a real person would actually answer. Ask for the warm intro. Sit beside the first users and watch them struggle. It is slow, it does not scale, and it is the single highest-leverage work you will do all year — because everything you learn in those hundred conversations becomes the thing that finally does scale.

Stop waiting for the world to arrive. Go get the first hundred by hand.

Your First 100 Users: Win Them One by One, by Hand