What shipping solo actually costs

7Sail is a solo studio. Several published titles, a website, a blog, a production routine that runs. And on the latest published game, a roguelite, the platform counters stayed flat.

I'm putting those numbers up front because everything else in this article depends on them. Not to complain: the industry produces thousands of invisible launches a year, and mine is nothing special. But you cannot talk about the real cost of shipping solo while implying things are going fine. They are going neither well nor badly: they are going the way they go for almost everyone, which is to say in silence. The subject is what you do with that silence.

Intuition says a studio that sells nothing spends nothing. That's wrong. The tools pile up, and almost all of them are subscriptions: the asset generation that stands in for the pixel artist I am not, the hosting for the site and the blog, the ecosystem around the engine. No single line item is unreasonable on its own; each one was chosen, compared, justified. But the total is recurring, and it lands every month against a revenue column that does not move.

It is not a problem of amounts, it is a problem of structure: a solo studio's costs are flows, its revenue is made of events. As long as no event arrives, the flow keeps running. A studio that doesn't count this literally does not know what each game costs it.

The heaviest line item shows up on no bank statement: time. Every week spent polishing a title almost nobody downloaded is a week I am not spending on the studio's flagship, a more ambitious app, currently on hold for lack of testers. The trade-off exists whether I look at it or not; looking at it is the only difference between arbitrating and enduring.

My answer is a blunt rule: WIP=2. Two projects in the finishing phase at most, at the same time, and not one more. It is not method for method's sake, it is a cap on attention debt. When a third project starts itching, the rule forces me to name the one that gets cut. It hurts exactly where it should.

Faced with flat counters, two reflexes present themselves, and they are symmetrical. Stop everything: the market has spoken, no point insisting. Or charge ahead: perseverance always pays off eventually, you just have to believe harder.

Both are comfortable because both excuse you from the same thing: measuring. Quitting without measuring means throwing away the work without knowing what it was worth. Continuing without measuring turns a project into a lottery where you buy a new ticket every week. The uncomfortable question, the one both extremes dodge, comes in two parts: what exactly am I measuring, and on what date do I decide?

My answer lives in boundaries written down before the emotion. The roguelite's release plan fits on one page and contains three things none of my previous plans had. A budget expressed in work sessions, not in features: when the sessions are spent, the plan stops. An explicit ending, written in black and white: at launch, the freed-up slot goes back to the flagship. And a traction threshold set before seeing the numbers, never adjusted afterwards to validate what I wanted to do anyway.

One studio rule completes the setup, the judge rule: a project is only finished once an outside, non-complicit person has given something in exchange. Money, time, a reasoned piece of feedback. Until that happens, the project is not finished, it is merely stopped.

Flat counters are not the record of a failure. It is a measurement, and a measurement only becomes a failure if you refuse to read it. Economic lucidity is not a manager's veneer laid over the craft: it is a studio skill on par with code, and it is learned the same way. By practicing on your real results. Even modest ones. Especially modest ones.