GifDrop

I just shipped GifDrop, a video-to-GIF converter for Android, built solo. The most counterintuitive decision I made is the one I keep thinking about.

GIF encoding on mobile is CPU-bound — there's no hardware acceleration for it. My first instinct was the obvious one: give users options. Let them pick high resolution, high frame rate, longer clips. More choice = better app, right?

Then I tested it. Higher settings meant conversions that dragged on for a long time on mid-range Android phones. The kind of wait where you assume the app has frozen. A terrible experience — and most of my target users are on exactly those mid-range devices.

So I did the opposite of what my instinct said: I capped it. Resolution, frame rate, and clip length all have deliberate ceilings. It felt wrong as a builder — you want to give people everything. But those caps are exactly what makes every conversion finish in seconds instead of minutes.

The reframe that stuck with me: the constraint isn't a limitation I'm hiding. It's the feature. "Fast, every time" is only possible because I removed the options that would make it slow.

Curious if others have hit this — a moment where giving users fewer choices made the product genuinely better? I suspect it's more common than we admit, especially solo where you can't out-engineer a bad default with a big team.