BUILDING IN KOREA
Manballers Devlog — Episode 09
Back to Zero — Simple Again, Sharp Again
The more features I added, the emptier it started to feel. So one night, I went back to the first question I’d ever asked myself.
The more features I added, the emptier it strangely started to feel. Screens kept multiplying, but the reason I’d started building this in the first place was getting harder to see clearly. One night, I went back to the very first question.
Why did I want to build this app to begin with?
01
Pulling Out the Original Reason
My father-in-law started barefoot walking for his health. My mother-in-law was always out hunting for a good trail to walk it on. Watching the two of them, a few specific frustrations kept coming up. Navigation apps only route you to the nearest subway station or bus stop — never the actual entrance to the trail. Travel somewhere new, and there’s no way to know which barefoot trail nearby is actually worth walking. Want to keep a record of the walk, and there’s no app built for that at all.
That was the original reason. And somewhere along the way, I’d also built a marketplace, a flash-meetup feature, and a finely tuned rank system. None of them were bad ideas — genuinely fun, genuinely plausible. The problem wasn’t any single feature. It was that, little by little, almost imperceptibly, the whole thing had drifted from the original purpose.
02
It Was Fine to Delete Almost All of It
I told Claude straight out — I’m fine deleting all of it, I just want this simple again. It wasn’t reluctance to lose what I’d already built. It mattered more that the app actually did the one thing it was meant to do.
The marketplace and flash-meetup feature didn’t get deleted outright — they got hidden. The code stayed exactly as it was; a single flag just switched them off on screen. If users grow and the need becomes clear later, they can be switched back on. Throwing away work I’d already put real time into felt wasteful, and rebuilding it later would cost that same time all over again.
The rank system stayed, but the metrics underneath it got trimmed. Originally it tracked three things — consecutive days walked, total days walked, total time walked. I couldn’t come up with a real reason “consecutive days” mattered, so it went. The walking/resting toggle, meanwhile, stopped being a status display and became an actual timer — a substitute for the smartwatch that most of this age group doesn’t own.
The problem wasn’t any single feature. It was that, little by little, the whole thing had drifted from the original purpose.
03
Dropping “Radius” Entirely
This was the biggest single decision. The idea of a “search radius in kilometers” was something users in their 50s and 60s found genuinely hard to reason about. It only produced confusion — “why isn’t this place showing up?” — with no real payoff.
I replaced it with two familiar, already-understood actions instead: “My Neighborhood” and “Search on the Map.” Curious about trails near Anyang? Just search Anyang on the map. No need to understand what a radius number even means.
I also rethought how pins get registered. The original method meant pinching, zooming, and dragging the map to line up an exact location — a level of fine motor precision I only later realized might be genuinely difficult for this age group. So the default became a single “Register Here” button, with a large on-screen prompt reading “Tap this at the entrance to the trail.” Manually adjusting the map pin got demoted to an exception, used only when the location comes out wrong.
I also reconsidered a feature that showed unrated restaurants and cafés as faded, semi-transparent markers on the neighborhood map. It had seemed like a good idea originally. In practice, I realized asking this age group to zoom in and out while distinguishing faded markers from solid ones wasn’t a fair ask. So it went. Now, only rated places show up as markers at all. Clearer won, every time.
04
Still Refining, Even Now
Even at the point of writing this, I’m still adjusting things. Tapping a region in the community screen should just change color like a tab, instead of navigating to a whole new screen. The map should show markers based on the currently visible screen area, not a kilometer radius. Markers should visually distinguish places I’ve rated from places others have rated.
None of it is finished. But the direction feels clear again, in a way it hadn’t for a while. And this time, instead of adding features on impulse the way I used to, I’ve started asking myself first — does this actually serve the original purpose? That question alone might be the biggest thing that’s changed over these last few days.