I shut down nine products. Here's why.
Eleven products since 2021, nine shut down, most within a month. I built what I thought would be fun, skipped the research, and barely marketed any of it.
I’ve built eleven products since 2021.
I shut down nine of them. Six lasted a month or less.
Looking back, the reason was almost always the same.
I built what I thought would be fun
Every one of them started with the same thought: wouldn’t it be fun if this existed?
I never asked whether anyone needed it. I did no research at all. I just started building.
Then I didn’t tell anyone
When I launched, I did almost no marketing. So users didn’t come.
When users didn’t come, my motivation dropped. When my motivation dropped, I stopped working on the product. Then I shut it down.
That happened again and again.
The one that was different
One New Friend is the only one that got real users. About 3,000 installs.
What I did differently was simple. I posted about it on Reddit, and most of the users came from there.
It’s still running today, in maintenance.
The one that lasted longest
My first product, Unbound Training, ran 229 days. That’s longer than the next two combined.
While I was building it, I was in Pioneer, an online accelerator anyone could join. It doesn’t exist anymore, but being part of it kept me going.
I wrote about this in One year later: I can’t stay motivated alone. Unbound Training is the proof.
Three I stopped before launch
Three of the nine never shipped.
- on-manage: I lost motivation.
- DreamFlow: it generated video with AI, and I couldn’t see how it would ever make a profit.
- MealDeal: I talked to people in the US on Reddit and other places. It didn’t solve a problem they needed solved right now.
MealDeal is the one time I actually did research. It told me to stop, and it was my shortest attempt: 11 days.
All nine
In the order they closed. The exact dates are on the products page.
- Unbound Training, online group fitness classes: 229 days
- prompt-career, a job board for prompt engineers: 19 days
- ship2post, GitHub releases turned into tweets: 114 days
- on-manage, AWS cost scheduling: 22 days
- anynomad, a digital nomad visa directory: 11 days
- SmartDigest, a daily email summarising X: 27 days
- DreamFlow, an AI dream journal: 53 days
- MealDeal, a meal planner built on grocery sales: 11 days
- GarageLess, mobile mechanic booking: 25 days
What I’d tell myself in 2021
Marketing matters more than building.
I knew how to build. I didn’t know how to get anyone to look. Nine products later, that’s the lesson.
If you’re building something and nobody’s showing up, I’d like to hear about it. I’m on X.