I built an event app and couldn't get anyone to use it

joinmyevent was a small event app with one good idea: guests could RSVP without signing up. The host makes an event, customizes an invite card, shares a link. Guests open it and respond, no account needed. That one bit of friction removed is genuinely nice, and most of the big players still make you sign up.

It was finished. And I mean actually finished: PHP and Slim on the backend, Postgres, payments wired up, 317 passing unit tests, end-to-end tests, blue-green deploys, spam protection, and the whole UI translated into five languages. It was live in production and it worked.

It reached almost nobody. This is the honest write-up of why.

The shape of it

I built the entire backend in a single day (December 12th, 2025). Auth, events, the no-signup guest responses, the subscription plumbing, all of it. Then over the next few weeks I added magic-link editing so guests who lose their cookie can still change their answer, plus-ones, custom questions, potluck mode, an invite-card maker.

Then it sat untouched for about three and a half months.

In June I came back and threw everything at marketing. The way I kicked that off is pretty me:

We have a strong product now. I want to market and improve further. Currently we only are missing users. What more features could we build? How could we market/monetize? Let's strategize. I think a fun idea would be to have a team of popular CEOs/CTOs discussing this using agents. Spawn 10 agents thinking like popular figures (Steve Jobs, ...) and have them discuss.

That became a recurring thing I did: convene ten agents role-playing founders and operators and have them argue a decision out before I built anything. It's genuinely good for pressure-testing, and twice it talked me out of a dumb move (repricing on eight data points, rebuilding billing before there was any traffic). About 160 commits went in over three weeks. And that sprint is where I got to watch, in real numbers, exactly how it fails.

The real numbers

I had analytics wired up from the start so I could see the funnel instead of guessing. Here is what it showed.

Organic search, over three months: 12 clicks. Average position 27.5, which is page three of Google. Dozens of SEO landing pages in five languages, and they basically never ranked.

So I paid for traffic. A two-week UK Google Ads run:

The top of the funnel was fine. People clicked, the cost per click was reasonable. Everything downstream fell off a cliff. Across the whole pricing system I'd built, the site took in zero euros, ever.

The number that actually stung came from the whole-site funnel: about 25 events got created, and across all of them roughly 28 invite links were ever opened. That's about one guest reached per event. A real party invite reaches ten to fifty people. Events were getting made and broadcast to nobody.

And the channels themselves closed one by one:

• Pinterest blocked every joinmyevent link as spam. The appeal form crashed when I tried to appeal.
• TikTok rate-limited me so hard I couldn't finish signing up.
• That left Instagram, where my account had two followers.

What people actually wanted

Here's the part that was useful to learn. When traffic did show up, it showed up for the invite card. My own read at the time:

People design the invite, and share that privately, but dont see a need in our event-tracking product.

The thing I thought was the product (tracking who's coming) was the part nobody wanted. The throwaway feature, a pretty card you can make in two minutes, was the part with pull.

I'll be fair to myself on one thing. A code review later found the card's share button was broken: it shipped the image without the RSVP link attached.

// public/js/card-editor.js
navigator.share({ files, title, text })   // no `url`, so the invite link never went along

So the one built-in way the thing was supposed to spread couldn't even fire. "Nobody wants this" is too strong. Closer to the truth: the loop was broken, and the demand that did show up was for the card. Both things were true, and neither one saved it.

What I actually learned

I've now done this enough times to name it. I build good things and I let them sit there. joinmyevent died because I have no audience and no foothold in the events world, and every channel to reach one was empty or blocked.

The slightly embarrassing part: I knew it while I was doing it. I wrote in my own strategy notes, mid-project, that it "has never gained traction for one reason: it was never distributed. Same failure mode as the founder's other projects." And then I kept shipping features anyway, because building is the part I'm good at and the part that feels like progress.

That pattern is the reason this site exists. Fix the habit underneath: build in the open, tell people honestly, grow an audience first so the next thing I make actually reaches someone.

If you're building something good and quietly hoping it'll find people on its own, it won't. I can tell you that one for free :)