Where to find your first 100 users (so you can measure anything at all)

Every guide to measuring product-market fit starts the same way: survey your engaged users. Nobody explains what to do when you don't have any yet.

This is the part that stalls more founders than a bad survey question ever does. You can't run a Sean Ellis survey on zero users, and a handful of friends-and-family signups won't give you a number worth trusting. Below 40 valid responses, any percentage you calculate is a rumor, not data. So before measurement, there's a different job: getting to 100 real users of the actual product, honestly enough that their PMF answers will mean something.

Why this stage gets skipped in most advice

Most PMF content assumes you're past this. It talks about survey wording, timing, and segmentation, all of which are useless without a base of people who've actually used what you built. The uncomfortable truth is that finding those first 100 is mostly not a marketing problem. It's a founder-hours problem, and it doesn't scale.

1. Your own network, worked directly

Start with people you can reach without a middleman: past colleagues, people in your industry, communities you were already a member of before you were a founder. This isn't "ask your mom to sign up." It's identifying the 20-30 people you know whose job or life actually intersects the problem you're solving, and asking each one individually, by name, why they might or might not want this.

Airbnb's founders didn't wait for traffic. Early on, Brian Chesky and Joe Gebbia flew to New York, went door to door to meet their hosts in person, and took the listing photos themselves because amateur photos were killing conversion. That one-by-one, unscalable push is credited with doubling their revenue in a month, not because it added many users, but because it taught the founders what was actually broken. Paul Graham wrote up the pattern across many YC startups in his essay "Do Things that Don't Scale": the founders who reach real usage fastest are almost always the ones manually recruiting and manually onboarding their first cohort, not the ones who built a funnel first.

2. Niche communities where your ICP already gathers

Somewhere online, the exact people who have your problem are already talking about it: a subreddit, a Slack or Discord community, a niche forum, an industry Facebook group. These groups convert far better than cold ads because the people in them have already self-selected into caring about the problem. Show up as a person contributing to the conversation, not a founder pitching a product, and the invitation to try what you're building lands very differently.

The trap here is spamming ten communities with the same pitch. Pick two or three where you can credibly participate, and treat the first users you get from them as people worth a follow-up conversation, not just a signup number.

3. One narrow beachhead, not a broad market

The fastest way to 100 real users is picking a market small enough that you can dominate it, not one so broad you can't find anyone in it. This is the logic behind the beachhead market framework: a startup that goes after "freelance designers who bill hourly and hate their current invoicing tool" will find and convince its first 100 faster than one going after "small businesses," even though the second market is larger on paper.

A narrow beachhead also pays off later. When you finally run the PMF survey, a tightly-defined group tends to answer more consistently, which means a cleaner signal instead of a blended score that hides what's actually working. That's the same reason segment-level PMF reads matter more than one blended number.

4. Founder-led onboarding, every single one

At this stage, every new user should get onboarded by a founder, personally. Not because it scales (it doesn't), but because it's the fastest way to learn what's confusing, what's missing, and whether the product actually solves the problem you think it does. Superhuman famously onboarded early users one-on-one over video calls before ever opening self-serve signup, which is part of why its early PMF survey answers were specific enough to act on. The full story is in how Superhuman went from 22% to 58%.

5. A waitlist with a real reason to say yes

A waitlist only works if there's a concrete reason to join it: early access, a founding-user price, direct input into the roadmap. "Sign up to be notified" converts poorly because it asks for an email in exchange for nothing. "Get access two weeks early and help shape what we build next" gives people a reason, and gives you a self-selected group who's already signaling real interest, not idle curiosity.

What "counts" as one of the 100

Not every signup counts toward your 40-user threshold. What matters for a PMF read is engaged usage: someone who has actually experienced the core value of the product, not someone who created an account and never returned. Chasing raw signups over engaged usage is how founders end up with 300 users and a PMF survey that says nothing, because most of those 300 never used the thing being measured. Better to work harder for 60 people who genuinely used the product than to have 600 who glanced at it once.