How to Get Your First 100 Users
Most side projects do not die of bad code; they die unseen, because distribution was left for after launch. The fix is structural: pick one channel before you build, build the product for that channel, then run the playbook that matches it. This guide covers the three playbooks, with a documented case study and the parts you can copy today.
Updated
One channel, chosen before you build
The first hundred users come from one channel, almost never from three at once. Spreading launch effort across every channel produces a weak showing in all of them; the projects that break through pick one and build the product so that channel can carry it. Search needs indexable pages. Product-led growth needs an invite worth sending. Social needs an artifact worth posting. None of those can be bolted on the week after launch, which is why the choice belongs in the first prompt.
This is also how our own prompt library is built. Every prompt carries the distribution mechanics for a named channel, and the coverage is measurable:
| Category | prompts |
|---|---|
| Product-led Growth | 67 (56 primary) |
| Email Marketing | 39 (5 primary) |
| B2B Sales | 34 (7 primary) |
| Organic SEO | 33 (14 primary) |
| Organic Socials | 20 (7 primary) |
| Community-led Growth | 19 (3 primary) |
Computed at build time from all 101 library prompts: total coverage counts a prompt once per channel it serves; "primary" is the channel it was designed for.
A documented case study, with sources
The strongest documented first-users story in this exact niche is cursor.directory, a free directory of Cursor rules. Its founder, Pontus Abrahamsson, has published the numbers himself: he built the first version in about three hours, launched it to the Hacker News front page, and reports the launch produced roughly 364,000 page views and a thousand GitHub stars. He later reported the site reaching $35k per month. Both figures are the founder's own, so treat them as self-reported rather than audited; the Hacker News launch and the GitHub stars are publicly visible.
Three things in that story are the transferable lesson, and none of them is "go viral":
- The artifact was launchable. A directory people could use in thirty seconds, free, with no signup. Launch platforms reward things a stranger can evaluate instantly.
- The code was public. Open-sourcing it turned the launch into GitHub stars, and the stars into an asset that keeps referring traffic years later.
- Launch traffic, not search, made the site. Despite a strong domain, its measured organic search traffic is a few hundred visits a month. The lesson cuts both ways: launches can dwarf SEO in this niche, and they decay, which is why the search path below still matters.
The three playbooks, honestly compared
| How it works | Speed | Failure mode | |
|---|---|---|---|
| Community-first | Two to four weeks of being genuinely useful in the two or three places your users already gather, then sharing what you built as a contribution rather than an ad. | Weeks. The most reliable of the three. | Posting your link before you have earned standing, which reads as spam and burns the community for good. |
| Search-first | Ship indexable pages that answer what your users already search for, from day one. The product IS the SEO surface. | Months to compound, then it keeps paying without launch-day effort. | Building a client-rendered app with nothing for a crawler to read, then wondering why nothing ranks. |
| Launch-first | Product Hunt, Hacker News, and the AI directories, with a launchable artifact: free, instant, self-explanatory. | Days, when it works. One strong spike plus permanent backlinks. | Launching a gated or unfinished product. Launch traffic does not forgive a signup wall in front of an empty room. |
The launch checklist that is actually checkable
Generic launch advice hides the useful part: where to submit. From our own competitor research (the referring-domain profiles of the sites that outrank us), these directories verifiably gave follow links to products in this niche: aitools.fyi, eliteai.tools, toolhunt.io, microlaunch.net, tinylaunch.com, launchtory.com and lobehub.com, alongside Product Hunt itself. That list is worth more than any launch-day tactic on this page, because links are the compounding asset a launch leaves behind.
Make the artifact free to evaluate
A stranger should get the point in thirty seconds without an account. If your core value sits behind signup, launch a free slice of it.
Prepare the one-liner and one honest screenshot per surface
Each directory wants a tagline, a description and an image. Write them once, well, and reuse.
Launch, then answer everything
The comments are the launch. Answer questions all day; each reply is content the next visitor reads.
Bank the links
The spike decays in days. The directory listings and any GitHub presence persist, and they are what search ranks you on later.
Build the channel into the first prompt
Whichever playbook fits, the product has to be built for it, and that decision lives in the prompt. The library tags every prompt with its channel, the generator asks for your channel before anything else, and product-channel fit is the reasoning behind the whole approach.
Frequently asked
How long does it take to get the first 100 users?↓
Should I launch on Product Hunt?↓
Do I need to build an audience first?↓
How does this connect to the prompts on this site?↓
Written by
Marco Kohns
Founder of ProtoBites - Venture Growth Studio
Growth PM at a Silicon Valley scale-up (a16z and General Catalyst backed), ex-Techstars where he consulted 13 early-stage startups, Reforge-trained. Every prompt on this site comes out of shipping ProtoBites' own portfolio products.