Notes on launching to crickets

When I first put PromptCraft online — the version with maybe forty prompts, the broken mobile layout, the machine-translated English — I tweeted about it once. Two people liked the tweet. One of them was my mom.

That's the whole launch story.

I'm telling it because the standard indie-launch narrative, the one you read on Twitter and IndieHackers, has a particular shape: founder works in stealth for six months, ships on Product Hunt, hits #1, gets 50,000 visitors on day one, and is off to the races. This narrative is not exactly *fictional* — there are people for whom it goes that way — but it's a vanishingly thin slice of what actually happens when you launch a product as a single person without a pre-existing audience.

The much more common story, and the one I want to write down because nobody else does, is: you ship, nobody notices, you keep going.

For the first three months PromptCraft existed publicly, the analytics dashboard was a depressing place. Twelve visitors one day, three the next, zero the day after. I'd refresh it and convince myself I was being delusional for thinking the work mattered. The voice in my head wasn't "you'll get there"; it was "maybe nobody actually wants this."

Here's the part nobody tells you: that voice is wrong, but not for the reason you think.

The reason it's wrong isn't that someone, somewhere, is about to notice and tell their friends. The reason it's wrong is that the question — "does anyone want this?" — is the wrong question to be asking when you're three months in. The right question is "do *you* still want this?" Because if the answer is yes, you'll keep working on it, and the work will keep getting better, and at some point — usually around month nine, if my data of one is generalizable — somebody who isn't your mom will show up.

The trap of optimizing for an audience that doesn't exist yet is that you start making decisions to please an imagined audience instead of the real audience, which at month three is you. You spend a week on a marketing page when the actual product still has rough edges. You agonize over your Product Hunt copy when nobody's coming from Product Hunt. You design for a launch that will never happen instead of the slow accretion of users who will arrive over time, one by one, mostly through search engines, mostly because the product is genuinely useful.

I wasn't wise enough at month three to know any of this. I just kept working on it because I was using it myself every day and couldn't stop noticing things that were broken or could be better. By the time the first stranger emailed me — month seven, I think — the product was three times as good as the version I'd shipped.

The first stranger's email said something like, "I've been using your prompt library for two weeks, it's really useful, thanks." I cried. Not because I was sad. Because somewhere in those four months of refresh-and-disappointment, I'd started to believe the work didn't matter. The email was proof that someone, somewhere, had been quietly using the thing without telling anyone, and the only reason I now knew was that one of them happened to be the kind of person who emails creators.

For every email I get, there are dozens of users who don't write. The Product Hunt-style launch wants to skip past this slow phase. It can't, really. The slow phase is where the work gets good enough that anything else can happen later.

So if you're three months into a project and nobody's noticed: that's normal. It's the most normal thing in the world. The thing to do isn't to launch louder. It's to keep shipping, keep using your own thing, keep fixing what bothers you, and trust that quiet adoption is the actual mechanism that grows indie products. Loud launches are theater. Quiet, consistent improvement is the real game.

PromptCraft is a small site by any reasonable measure. Most months its numbers don't impress anyone. But it's been growing, slowly, for nine months — and the slope of that growth, week over week, has barely changed. That's what compounding looks like when you're the one doing it. It's invisible until it isn't.

Don't optimize for the launch. Optimize for the version of the product that exists nine months after the launch nobody noticed.