Every guide on how to get first 100 users seems to start from a cheat code: "I tweeted it to my 30k followers." Great. You have 43 followers and 12 of them are bots. The good news is that the no-audience playbook isn't a worse version of the audience playbook — it's a different one, and it's the one almost every builder actually runs, because almost nobody has an audience the 1st time. The 1st 100 users don't come from reach. They come from showing up, individually findable, in places where your exact problem is already being discussed. This post lays out the 5 channels that work from 0, how to run them without being the spam everyone hates, and why picking 1 or 2 beats trying all 5.
Communities: earn the right to post the link
Reddit, Hacker News, and niche Discords are where no-audience launches actually happen — and where most of them die, because builders treat communities as billboards. The failure mode is always the same: account created Tuesday, link posted Wednesday, banned Thursday. Communities convert in proportion to how obviously you belong there, and belonging is checkable in 5 seconds via your post history.
The non-spammy version: spend 2-3 weeks being genuinely useful in the 2-3 communities where your users complain about the problem you solve. Answer questions. Share what you learned building, including the failures — build-logs consistently outperform launch posts. When you do share the product, share it as a story with numbers ("I built X because Y kept happening; here's what 30 days of building taught me") and put the link where the community's norms allow, sometimes that's a comment, not the post. On HN, a Show HN with an honest, technical 1st comment from you is the format. Expect most posts to do nothing; the 1 that lands can be worth 20-50 signups, hedged, sometimes far more. That's the channel: low floor, spiky ceiling, compounding trust.
Directories and launch platforms: cheap, stackable, mostly small
Product Hunt, the Indie Hackers product directory, AlternativeTo, free-tool listings, awesome-lists on GitHub — individually, each might send you a trickle, commonly single-digit signups, though outliers happen. The move is to treat them as a batch: 1 afternoon, 10-20 listings, done. You're buying 3 things — a handful of users, backlinks that feed the SEO long game, and the social proof of existing in more than 1 place when a prospect googles you.
Launch platforms like Product Hunt deserve 1 honest caveat: a big PH day is mostly traffic from other makers, which is great if makers are your market and a vanity spike if they're not. Take the spike, screenshot the badge, and don't confuse it with distribution. Real distribution is the thing you should've been thinking about even before the build — we've made that argument in full in distribution before you build.
Cold-but-relevant outreach: the channel nobody wants and everybody needs
Cold outreach has the worst reputation and, from 0, some of the best unit economics — because at this stage you don't need scale, you need 100 humans, and 100 is a hand-findable number. The qualifier is "relevant": not a purchased list, but people who've publicly displayed the problem. They posted the complaint on Reddit, asked the question on a forum, reviewed a competitor 2 stars, tweeted the workaround.
The message that works is short, specific, and embarrassingly honest: I saw your post about X. I built a thing that does Y. It's new and rough — want a free account in exchange for blunt feedback? No sequence, no "just bumping this." Reply rates on genuinely relevant cold messages can plausibly hit somewhere around 10-30% — hedged, it swings hard with targeting — versus the ~1% of spray-and-pray. 10 sends a day is a habit, not a job, and as a bonus the replies are user research you'd otherwise pay for: the 1st 20 conversations will rewrite your landing page.
SEO and small paid tests: the long game and the thermometer
SEO from a 0-authority domain is a 6-12 month bet, roughly and depending on the niche — so it's the worst channel for your 1st 100 users and one of the best for your next 10,000. Start it now, expect nothing soon: a handful of pages targeting the exact long-tail questions your users ask, comparison pages, an honest changelog. The community posts and directory listings above double as your 1st backlinks, which is a tidy bit of channel synergy.
Small paid tests are the opposite shape: instant, expensive, and clarifying. $100-200 on tightly-targeted Reddit or Google ads — hedged, your numbers will vary — won't buy you 100 users at sane prices, and that's fine, because you're not buying users, you're buying data. Which headline gets clicked, which audience converts, what a visitor costs. Paid is a thermometer at this stage, not a fuel line. If a cold $150 test converts strangers at all, you've learned something most unlaunched projects never learn; if it converts nobody, that's cheaper to know now. Either way, instrument the funnel before you spend — know your funnel performance covers what to measure — because paid traffic into an unmeasured funnel is just buying mysteries.
Pick 1 or 2 channels and go embarrassingly deep
Here's the part that separates the builders who get to 100 from the ones who get to 11: depth beats breadth, and it isn't close. 5 channels at 20% effort produces 5 rounding errors. 1 channel at 100% — daily presence in 1 subreddit, or 10 relevant cold messages every weekday — produces compounding returns: recognition, trust, referrals, the moderator who knows you.
Choose by fit, not fashion. Where do your users already gather, and which channel matches your product's motion? Developer tool: HN, Discords, GitHub. Prosumer app: Reddit, directories, SEO. Niche B2B: cold outreach, almost exclusively. This is distribution-product fit — the match between what you built and how it's findable — and a mediocre product with a fitting channel routinely outperforms a great product shouting into the wrong room.
The honest catch: all of this is hours, every week, for months, and it's exactly the work that leaves projects rotting in a tab — the building was the fun part, and this is the other part. That gap is what LaunchBuddy exists for. It's a launch studio: submit your unlaunched project, and if it's picked, we build it onto the harness, launch it live, and operate this entire playbook — channels, funnel, iteration — as the operating work it actually is. You keep ownership: flat fee or a rev-share where you'd keep roughly 70% rising toward 90% after build costs recoup (placeholders, hedged until signed), and you can kill or port out anytime.
Want to know if your project is worth 100 users' attention before you grind for them? Free, honest assessment at launchbuddy.app — if it's a no, you get the why. 60 seconds to submit. Cheaper than 1 more month of the tab staying open.