How do you get your first 100 users for a startup?
Your first 100 users will almost certainly come from manual, unscalable effort—direct outreach, personal recruitment, and hands-on onboarding—not from a launch, a press hit, or a partnership deal. The goal at this stage isn't efficiency; it's learning what makes even one person genuinely need your product. Once you have that, you recruit the next person the same way, and then the next.
Forget the launch—recruit users one at a time
Most first-time founders overinvest in launch strategy and underinvest in direct user recruitment. A launch gives you a momentary spike of curiosity-driven signups who mostly churn; a personally recruited user who really needs what you built gives you a relationship, feedback, and a potential evangelist. Paul Graham's observation about launches is blunt: what actually determines your trajectory a few months later is how happy those early users are, not how many showed up on day one.
The practical implication is to start with people you can actually reach—your professional network, former colleagues, communities you're already part of, anyone who has explicitly complained about the problem you're solving. Send individual messages, not a mass email. Explain exactly why you thought of them specifically. This feels slow, but it compounds: a satisfied user in a tight-knit professional community will tell two or three peers without any prompting.
Resist the temptation to wait until the product is 'ready' for a big moment. The product will never feel ready, and the users you recruit manually during the rough early phase are exactly the ones whose feedback will make it worth launching later.
Go to your users physically if you have to
The Airbnb story is the canonical example of this principle in action. Graham describes how the founders flew to their early hosts' locations, photographed listings themselves, and stayed with users to understand the experience firsthand. About 30 days of that kind of direct, in-person engagement—unglamorous and completely unscalable—was the margin between survival and failure for the company.
For most B2C products, 'going to users' means showing up in the places they already gather: subreddits, Slack communities, Discord servers, industry meetups, local Facebook groups. Don't post a generic announcement. Engage with conversations where people are already expressing the problem, and reach out directly to individuals who seem to be a strong fit. For B2B products, this often means physically visiting an early customer's office to watch them use your product in context—you will learn things about their workflow that no survey or analytics tool would ever surface.
The founders who grow fastest early are usually the ones willing to do the most personally uncomfortable things to get close to users. Cold DMs, showing up to niche conferences, asking friends-of-friends for introductions—none of it is scalable, all of it works.
Start in a deliberately small, specific market
One of the most counterintuitive moves for getting your first 100 users is to narrow your target market until it feels almost too small. Facebook launched exclusively for Harvard students. The constraint wasn't a limitation—it was the reason early users felt the product was genuinely made for them, which drove the density of adoption needed to make a social product work at all.
For your startup, identify the single most specific category of person who has the sharpest version of the problem you solve. Not 'small business owners'—'independent bookkeepers who work with e-commerce clients on Shopify.' That level of specificity lets you find them in concentrated communities, speak their exact language, and build features that fit their workflow so precisely they can't imagine going back to whatever they were doing before.
The tactic of treating one early user almost like a consulting client—building closely around their specific needs—is something Graham explicitly recommends for B2B startups. Pick one person or company that really needs what you're making, serve them so completely that they become a reference case, and use that reference to open doors to the next ten users who look just like them. This is not a detour from building a scalable product; it's the fastest path to understanding what that product actually needs to be.
Make your first users unreasonably happy
The standard for customer attention at large companies—automated support, generic onboarding, impersonal communication—is not a model for early-stage startups. It's actually the opposite of what works. As Graham notes, the founders who obsess over delighting individual early users have never once been led astray by it; the downside risk of being too attentive to users is essentially zero.
In practice this means: onboard every early user personally, via a call or screen share if at all possible. Message them a few days after they sign up to ask what's confusing. When something breaks, fix it and tell them personally what happened and what you changed. Send a note when you ship a feature they asked for. None of this takes enormous time at sub-100-user scale, and it generates a quality of feedback and loyalty that paid acquisition can't replicate.
Founders often resist this level of attention because it doesn't feel professional—it feels desperate. That instinct is wrong. Garry Tan has pointed out that new founders sometimes imitate the indifference of big companies because it seems more 'serious,' but that indifference is actually a bug in large organizations, not a feature worth copying. Your smallness is an asset: use it to give every early user an experience so good they feel compelled to tell someone about it.
What not to rely on: launches, partnerships, and paid ads
Three tactics absorb enormous founder energy at the early stage while reliably underdelivering: a big public launch, a partnership with a larger company, and paid acquisition. Each feels like a shortcut to user volume; each almost always disappoints.
Partnerships with established companies are particularly seductive because they seem to outsource the hard work of distribution. But partnerships require the larger company to prioritize your product, which never happens when you're small. Six months of negotiations, a signed agreement, and then—nothing, because your contact left or the priority shifted. The same hours spent doing direct outreach to individual potential users would have gotten you further.
Paid acquisition can work eventually, but before you understand what makes your best users convert and retain, you're paying to learn lessons you could have learned for free by talking to people. Spend that budget on travel to meet users in person, or on events in communities where your target users concentrate, before you spend it on ads. The signal you get from a human conversation is worth far more than the aggregate data from a thousand anonymous clicks when you're still figuring out product-market fit.
“Almost all startups are fragile initially… The big danger is that you'll dismiss your startup yourself.”
— Paul Graham, source
The one thing to do
Pick five specific people who have the exact problem you solve, contact them individually today with a message that shows you understand their situation, and onboard whoever responds as if they were your only customer.
Frequently asked questions
Should I launch on Product Hunt to get my first users?
A Product Hunt launch can generate a short burst of signups, but those users are often product enthusiasts rather than your actual target customer. Use it as a supplement to direct recruiting, not a replacement—and only once you've validated the product with at least a handful of manually recruited users who genuinely need it.
How do I find people to recruit as early users if I don't have a relevant network?
Go where the problem lives: subreddits, industry Slack groups, LinkedIn searches, niche newsletters, and professional associations. Post helpfully in those communities before you pitch anything. When you do reach out, be specific about why their situation matches what you're building—generic cold outreach almost never works, but a precise message that clearly understands someone's problem gets replies.
How many early users do I need before I stop doing things manually?
Don't think of it as a threshold to cross. The manual, high-touch approach should continue until the volume of users makes it physically impossible—and even then, you should systematize the attentiveness rather than eliminate it. Many founders stop too early, before they've extracted the learning that only direct user relationships provide.
What if my target market is enterprises and I can't just DM potential users?
Cold outreach and warm introductions through your network are your primary tools. If you have no connections in the industry you're targeting, that's worth reflecting on—domain expertise and relationships matter enormously for enterprise sales. Start with the smallest enterprises where you can reach decision-makers directly, and use early customers as references to open doors at larger ones.
Sources
- Billionaires Build — Paul Graham
- Do Things that Don't Scale — Paul Graham