Twenty-Eight Signups from a Story, Not a Launch

A few weeks ago I wrote about launching PuraLetra on Product Hunt and getting exactly zero signups. I reframed the failure at the end of that essay. The problem wasn't that nobody wanted a clean article reader. The problem was that I hadn't figured out how to get in front of the people who might.

Turns out, the people who might were on Threads.

I posted about PuraLetra on Threads. Nothing elaborate. Just this:

"I built a reading app after Pocket shut down. Claude Code was my partner. Six months of work. I use it every day. Launched it on Product Hunt. Nobody came."

That was it. Within a day or two, the post had more engagement than anything I'd posted on any platform about PuraLetra. People replied. People signed up. Not thousands. Not hundreds. But real people, creating accounts, saving articles, using the thing I'd built.

Jonathan Bruck — an early advisor to Pocket — commented on my post. He talked about how Pocket's secret to distribution was getting journalists to use it, building the first share sheet SDK that every app relied on before iOS made its own. Then he landed on the same point I'd been circling: it's easier to build now, but harder to reach users.

Someone who helped build the thing that inspired my thing was telling me, publicly, that the problem I was facing was the same problem they'd solved, not with a better product, but with a better path to the people who needed it.

The most common thread in the replies wasn't about PuraLetra itself. It was about AI. People used my post to confirm something they already believed: that AI is making it easy for anyone to build an app, but the hard part, finding users, selling, distributing, is what separates a side project from a product. That's the wall these "vibe-coded" apps are going to hit.

I get the argument. And honestly, they're not entirely wrong. Distribution is the hard part. That's the whole point of this essay.

But I want to be precise about something, because it matters to me. I didn't vibe-code PuraLetra. I didn't describe an app in a prompt and wait for something to appear. I used Claude Code as a tool to augment what I already know how to do. I've been a software engineer for over two decades. I understand databases, architecture, tradeoffs. What Claude Code gave me was speed and reach — the ability to move through unfamiliar frontend territory faster, to stay in flow instead of stopping to search Stack Overflow every ten minutes, to build at the pace of my ideas instead of the pace of my typing.

There's a difference between using AI to skip the thinking and using AI to accelerate it. I did the thinking. I made the architectural decisions. I chose the tradeoffs. Claude Code was the best collaborator I've ever had, but it was a collaborator, not the architect.

That distinction might not matter to the people scrolling past my post. But it matters to me. And if you're an engineer reading this, it probably matters to you too.

Here's what I actually learned from the contrast between Product Hunt and Threads.

Product Hunt felt like the correct place to launch a product. It's where you're supposed to go. There's a whole ritual around it — the timing, the tagline, the first-day push. I followed it. I got nothing. Not because PuraLetra was wrong for that audience, but because I had no audience there. I was a stranger walking into a crowded room and hoping someone would notice.

Threads worked because I wasn't launching. I was telling a story in three lines. I built something. I use it every day. Nobody came. That honesty resonated more than any tagline. And the Pocket reference did most of the heavy lifting — it gave people an instant frame of reference. They didn't need to understand what a "clean article reader" was. They already missed Pocket. They already knew the gap.

Distribution isn't a step you bolt on after building. It's not a launch day. It's the accumulation of small moments where you show up in the right room and describe the right problem to people who already feel it. I'd been showing up on Threads for weeks before that post took off. Not strategically. Just consistently. And when the right post landed, there was enough context around it for people to take it seriously.

I think about this in the context of everything I'm trying to do with this newsletter. I spent months building PuraLetra. The product was ready long before anyone used it. What wasn't ready was the path between the product and the people who needed it.

In the last essay I wrote that I was shifting from "nobody wants this" to "I haven't found the right people yet." The Threads response confirmed the reframe. The people exist. The need is real. I just had to stop performing a launch and start having a conversation.

I don't have a grand theory about distribution. I have one data point. But it's the first data point that actually moved the needle, and I'm paying attention to what it's telling me.

If you want to try PuraLetra, it's at puraletra.com. And if you want to follow this experiment as it unfolds, I'll be here next week.