How successful startups found their first users

The thing nobody tells you about early traction is that the companies you admire didn't get there with clever funnels or growth hacks. They got there by doing things that were embarrassingly manual, deeply personal, and completely impossible to repeat at scale.

Paul Graham wrote about this in his essay "Do Things That Don't Scale" — but most founders read it once and move on. These stories are what that advice actually looks like in practice.

Airbnb: Go to your users' homes

In 2008, Airbnb's listings were bad. Bad photos, low conversions, struggling revenue. The team figured out why: people couldn't tell what they were booking from the grainy phone shots hosts were uploading.

So Brian Chesky and the team flew to New York, knocked on hosts' doors, borrowed a camera, and photographed their apartments themselves. Then they went to the next host. Then the next.

Revenue doubled. Not from a product change or an ad campaign — from two founders doing something no investor would call a strategy. That experiment gave them the insight that eventually became a professional photography program that scaled.

They also personally hosted their early guests and met every user they could in the first months. That's how they found out what was broken before they tried to grow.

Stripe: Set it up for you right now

When a developer agreed to try Stripe in the early days, Patrick and John Collison didn't send them a link. They asked: "Can I set it up on your laptop right now?"

They'd sit down, open the developer's machine, and integrate Stripe on the spot. Paul Graham named this the "Collison installation" because it was so reliably effective that he wanted every founder to copy it.

The instinct most founders have is to optimize for signups. The Collisons optimized for activation. Those are different problems, and solving the second one first is almost always the right call.

Reddit: Two founders, hundreds of accounts

Reddit launched in 2005 and felt alive from day one. It wasn't. For the first several weeks, roughly 99% of posts were Steve Huffman and Alexis Ohanian submitting links under fake usernames.

"We basically faked our first users," Ohanian said later. "We set the tone and made the site feel alive."

This is the cold start problem in its purest form. Social products with no users have no value. The only way to break that loop is to manufacture the initial critical mass yourself — not by cheating, but by being the user until real users show up and take over.

For months, 80% of Reddit's submitted links came from its two founders. Then the community slowly began to carry itself.

Dropbox: Build the waitlist before the product

Drew Houston made a three-minute demo video of Dropbox before the product was ready to ship. He posted it to Hacker News. The waitlist went from 5,000 to 75,000 people overnight.

The product didn't exist in a usable form. But the problem was real, the demo made it concrete, and the waitlist told Houston he was building something people would actually pay for before he over-invested in building the wrong version.

Later, Dropbox's referral program gave users 128 MB of storage for each friend they referred. It sent 2.8 million invites in 30 days. But that only worked because the early users were already passionate — you can't bolt virality onto something people don't genuinely want to share.

Tinder: Campus by campus, sorority by sorority

Before Tinder had 5,000 users, Whitney Wolfe (then Head of Marketing) drove to college campuses across the country and went directly to sorority chapters. She'd present the app at chapter meetings, explain what it was, and have every woman in the room install it on the spot.

Then she'd cross the street to the fraternity house and tell the guys: "There are 150 women on this app at your school right now."

When she returned from that tour, Tinder had grown from 5,000 to 15,000 users. No ads. No algorithm. Just a person going room by room.

The insight here was about density. Dating apps are useless if there are no matches nearby. By seeding them campus by campus, Tinder manufactured local density before trying to grow nationally.

DoorDash: One neighborhood at a time

DoorDash's early team didn't try to launch across a city. They targeted Stanford's campus, knocked on doors of local restaurants, signed them up manually, and pitched the service person-to-person to students — specifically sororities and fraternities.

The goal was density in a single ZIP code, not breadth across a metro. A product that works perfectly in one neighborhood teaches you far more than one that works passably across an entire city.

That granular understanding of delivery economics and customer behavior is what let DoorDash see something its competitors missed: suburbs had better unit economics than dense urban cores. While other players burned cash competing in cities, DoorDash quietly dominated the neighborhoods nobody else was paying attention to.

Pinterest: Opening it on strangers' computers

In Pinterest's early days, Ben Silbermann promoted the platform to friends and family, hosted local events to get people using it, and reached out to bloggers to cover it. But the detail that sticks: he reportedly went into stores and opened Pinterest on the display computers and iPads so passersby would encounter it.

It's the kind of thing that sounds absurd as a growth strategy. It's also the kind of thing that reflects a founder who is willing to do literally anything to get people to see what they've built. Pinterest eventually grew to tens of millions of users, but it spent two years being cultivated almost entirely by hand.

Product Hunt: An email list with 20 subscribers

Ryan Hoover launched Product Hunt in November 2013 as a Linkydink email list — a simple tool for group email curation. He spent about 30 minutes setting it up one morning, invited a handful of friends to contribute, and shared it on Twitter and Quibb.

Within a week, hundreds of people had subscribed. Within 20 days, it had 2,000 subscribers. That was enough signal to justify building a real product. Hoover and a friend built the first version of the Product Hunt website over Thanksgiving break.

The whole thing started because Hoover personally had the problem: he wanted a better way to find new products being built. The email list was him solving it for himself and a few dozen friends. It turned out a lot of people had the same problem.

Superhuman: A 30-minute call with every single user

For years, every person who joined Superhuman — the email client — had to complete a 30-minute one-on-one onboarding call before they could use the product. An onboarding specialist (or, early on, Rahul Vohra himself) would configure the app to their workflow, import their email, teach them the keyboard shortcuts, and get them to inbox zero during the call.

Superhuman had a waitlist of 550,000 people at one point. Every one of them was waiting for that 30-minute session.

Rahul personally onboarded hundreds of early users — including the co-founder of Instagram and Justin Bieber's manager. That direct contact shaped every product decision in the early years.

The result: users who completed onboarding became evangelists. Word of mouth drove more demand than any paid channel. The concierge model wasn't a bottleneck — it was the growth engine.

The patterns underneath all of this

These aren't random stories. The same things repeat:

Go to them, don't wait. Airbnb flew to New York. Stripe sat next to developers. Tinder drove to campuses. DoorDash knocked on restaurant doors. Early users rarely come to you — you have to find them where they already are.

Own a small thing completely before going broad. Facebook launched only at Harvard. Tinder launched campus by campus. DoorDash launched in one ZIP code. Uber launched in San Francisco only. Depth in a small cohort beats breadth across a large one every time.

Seed the community yourself until it can carry itself. Reddit's founders were 80–99% of the activity for months. Product Hunt's first email list was Hoover and a handful of friends. You are the first user, the second user, and probably the hundredth.

Manufacture the first-use experience by hand. Superhuman's 30-minute call. Stripe's Collison installation. Airbnb's professional photos. The unscalable thing you do for the first hundred users is what teaches you what "great" actually feels like — and that's the template everything else is built from.

Treat early users as a research project, not a revenue target. Every company on this list was learning from their first users at close range. The founders who sat with users, watched them struggle, and asked "why" built things that grew. The ones who tracked them from a dashboard usually didn't.

There is no shortcut that replaces any of this. The companies that look like overnight successes from the outside spent months doing things that looked embarrassingly small from the inside. That's not a bug in the process — it's the process.