I had a clean exit strategy going in. Ship the minimum, see if anyone used it, then decide.
One day after paying Kudoboard $5.99 for a group greeting card at work, I had my own version live at ricoordonio.com/kudos. Text kudos, a shared link, auto-deletes after 30 days. MVP. Done.
Then I added a Giphy picker.
Then I rebuilt the card rendering to support embedded videos. Then themes: Bold first, then Cozy, then Elegant. An Outdoor theme that evokes the place I work without using the name. A Blocks theme meant to feel like LEGO without touching the IP. Vintage. Cyber. A Comic theme with actual ink-style card treatments.
Seven themes in four days.
At some point I also wired up a /kudos/changelog page that generates itself from git history. The commits already read like release notes: “Kudos: add Cyber, Vintage, Comic themes with full card treatments.” I just needed a page to surface them. The git log became documentation without me writing any documentation.
And then I removed Kudos from the ricoordonio.com header navigation entirely.
That last move is the tell. Not the themes. Not the changelog. The moment I decided it shouldn’t be anchored to the parent site anymore, that was the signal. The project was starting to have its own identity. It’s heading toward its own domain, and it’ll take a proper AWS-native backend with it when it goes.
I said I had a clean exit strategy. What I actually had was a project finding out what it wanted to be.
Don’t get me wrong, this is textbook scope creep. I have MVP rules. Minimal viable, then decide. I broke them. But there’s a useful distinction between scope creep and product discovery, and the way you tell them apart is whether you’re padding a weak core or whether the core keeps pulling you back.
Every theme came from the previous one suggesting a direction. Bold’s editorial card design made me think: what’s the opposite of editorial? That became Cozy. All of the themes were looking pretty modern, what about something older? Vintage. The git history is a record of one idea following the next.
I used to think compulsive building was a discipline problem. A failure to scope.
It can be. But sometimes what you’re calling scope creep is a signal you’ve stumbled into something worth building for real.
If you’re still adding features two weeks after shipping the MVP, don’t ask why you haven’t stopped. Ask why you ever thought you would.