Three games. Three launches on itch.io. Three flops — depending on how you measure "flop," which I'll come to. The instinct after a quiet launch is to blame the game. Sometimes that's right. More often it's not, and you ship the next game with the same blind spots and watch it tank for the same reasons.
So: here's what actually went wrong. With names anonymised because the games are still up and I'd rather not bury them further.
Game one — the screenshot problem
A 2D action platformer. Six months of work. We were proud of the movement; the playtest group loved the movement; everyone we'd shown it to said it felt great.
The store page launched with five screenshots of static gameplay. No trailer. The most-played frame in our best level — a chase sequence with the player on a collapsing bridge — was reduced to a single still photograph. The game's biggest selling point was its motion, and we'd represented it as a series of stills.
What we should have done: a thirty-second silent loop of the chase sequence, embedded above the fold, autoplaying. Itch.io supports this. Most discovery-page hovers are silent — sound off, sound on later. We optimised for the wrong frame.
Game two — the title problem
A puzzle game with a clever core mechanic. Title was a made-up word — playful, distinctive, completely meaningless to anyone scrolling.
Itch's discovery surfaces are a wall of titles and thumbnails. People scan, they don't read. A made-up title forces a reader to think, and at scan speed they don't. They scroll past.
The brutal fact: game title is the second-biggest signal you give the discovery algorithm and the player, after the thumbnail. We renamed the game six months later — incorporated the mechanic's actual name into the title — and the click-through doubled overnight from a base of "almost nothing." The game itself hadn't changed.
If you can't summarise what your game is in your title and thumbnail, that's the lesson, not the game.
Game three — the audience-building problem
The biggest project. Eighteen months. A vertical-slice demo released six months before the full launch. The plan was: build audience with the demo, convert at launch.
The demo got around three hundred downloads. Of those, maybe forty wishlisted on Steam. At launch, the conversion rate from wishlist was strong — but the absolute number was forty. That isn't an audience. That's a focus group.
The mistake wasn't releasing a demo. The mistake was treating the demo as the audience-building. Audience building happens upstream — devlogs, tweets, screenshots, gifs, a presence in the genre's communities, a mailing list. The demo is the closer, not the prospector. We had no upstream, so the demo had nobody to close.
The next launch we ran a six-month devlog campaign starting from prototype. By the time the demo dropped, there were two thousand people who already knew the game existed. The demo still got around three hundred downloads — but the wishlist conversion was an order of magnitude higher because the people clicking already cared.
What "flop" actually means
Worth saying: none of these three games made nothing. They paid for themselves in single-digit thousands of pounds. They weren't business failures so much as growth failures — the next game was supposed to be twice as big as the previous and instead it was the same size. The audience wasn't compounding.
That's the metric that matters for indies: not "did the game succeed" but "did your audience grow by enough that the next game has a bigger floor than this one had a ceiling?" Three games in, ours hadn't. That was the real flop.
What changed
Trailers above the fold. Titles that name the thing. Devlogs from prototype, not from demo. Communities first, demos second, launches third. The next game shipped this year and the wishlist count on launch day was bigger than the lifetime players of the first three combined.
None of this is news to anyone who's been in indie marketing a while. But these aren't things you can hear once and absorb — you have to live them, watch a launch fail because of them, and feel it in your bank account. That's why they stick.
If you're three weeks from shipping and your trailer is a series of stills, your title is a made-up word, and your audience is whoever happens to be on the front page of itch that morning — postpone. Three weeks of fixing those three things will return more than three months of feature work.