What a Small Game Studio Spends in Its First Year
Someone in our Discord asked what it costs to start a studio, and the answers he got ranged from nothing to thirty thousand dollars. Both numbers are technically true and neither one helped him.

Someone in our Discord asked what it costs to start a studio, and the answers he got ranged from nothing to thirty thousand dollars. Both numbers are technically true and neither one helped him. So I went back through our own records from the first twelve months of Fox Studios and added up every line that left a bank account.
The total was 1,042 dollars, spread across three people, two released games and one that we cancelled. Most of it went to things nobody warns you about, and the expensive items I expected to pay for turned out to be free. Here is the breakdown, with the mistakes left in.
Where the money actually went
| Item | What we paid | Was it worth it |
|---|---|---|
| Steam Direct fee, two games | 200 | Yes, recouped after 1,000 dollars of sales each |
| Capsule and store art, commissioned | 340 | Yes, the single best line on this list |
| Music, two licensed tracks plus one commission | 185 | The commission yes, the licensed tracks no |
| Sound effects pack | 22 | Yes, still using it |
| Aseprite, two seats | 40 | Yes |
| Domain and email, one year | 56 | Yes, hosting was free |
| Font licence for the logo | 35 | Debatable, a free face would have worked |
| Paid ads experiment | 120 | No, and I explain why below |
| Hardware, one secondhand controller for testing | 44 | Yes |
Nothing on that list is an engine, a subscription or an office. Godot cost us zero. Our site runs on free static hosting. The task board, the version control, the build scripts and the bug tracker were all free tiers that we have still not outgrown at three people.
What surprised me was how much of the real cost sat in two categories: art that faces the store, and audio. Those are the parts a programmer cannot fake, and they are the parts players judge in the first four seconds.
The free tools that replaced things I assumed we would buy
Before we started I had a spreadsheet of software I thought a studio needed. Almost all of it had a free answer that was good enough to ship with.
- Engine. Godot, free, no revenue share. Unity's personal tier is also free under a revenue threshold, which is well above where most first year studios land.
- Video editing for the trailer. DaVinci Resolve, free version. We cut a 68 second trailer in it. The paid Studio version is around 295 dollars and we have never needed it.
- Audio editing. Audacity for cleanup, free. Reaper is 60 dollars for a personal licence if you start writing your own music, which is the one paid tool I would add next.
- Tilemaps and levels. Tiled, free, and it exports in a format Godot reads directly.
- Builds and uploads. The itch.io command line tool, free, and it turns a release into one command from the terminal.
The pattern here held for the whole year. If a tool was for us, there was a free version that worked. If a tool produced something a player would see or hear, paying was usually the right call.
Why we would not spend that 120 dollars on ads again
The advice you hear constantly is that a small studio needs a marketing budget. We tested it. We put 120 dollars into a short campaign on one social platform in the two weeks before a launch, targeted at people who followed similar games.
It produced 31 clicks to the store page and, as far as we can tell from the wishlist graph, four wishlists. Four. In the same two weeks, one devlog post that cost nothing but an evening brought in about 190 wishlists, and a single comment I left on another developer's post brought in more than the entire paid campaign.
I am not claiming ads never work. At a budget of 120 dollars you cannot learn anything, because you cannot run the test twice with a change in between. The honest version is that paid promotion is a tool for teams that already know their cost per wishlist, and a first year studio does not. Spend that money on capsule art instead, which keeps working on every visit to your page forever. Our notes on getting Steam wishlists before launch cover the free routes that did work for us.
The costs that are not money
Our biggest expense in year one was a cancelled project, and it never showed up in a bank statement. We spent about five months on a game that we shelved at 60 percent done because the core loop was not fun and we had known that since month two.
At even a modest valuation of our own time, that dwarfs the 1,042 dollars by a factor of twenty. The lesson we took was not to stop prototyping. It was to set a date where we ask one question out loud: would any of us choose to play this tonight. If the answer is no twice in a row, the project stops. That is the cheapest cost control we have found, and it is the reason we keep our scope small on purpose.
What we skipped on purpose
We did not register a company in year one, and I would make the same choice again at our size. Filing fees in our case would have run between 90 and 400 dollars depending on structure, plus an accountant we did not need yet, to protect revenue that had not arrived. We used personal accounts, kept a plain spreadsheet of every transaction, and agreed the split in writing first. The revenue split agreement post describes what we wrote down.
Once a project clears roughly four figures a month, or the moment you take money from anyone outside the team, the calculation flips and you should talk to someone qualified in your country. We are not there and I would rather say that plainly than pretend otherwise.
The other skips: no asset store shopping sprees, no paid Discord boosts, no conference travel, no paid press release service. The last one gets pitched to small studios constantly for 200 to 500 dollars. Every journalist I have asked says they ignore those and read direct emails.
A budget you could copy next month
If I were starting again with one small game and three people, I would plan for about 500 dollars and no more. Roughly 100 for the Steam Direct fee, 250 for capsule and store art from an artist whose work you already like, 100 for music or a commissioned theme, and 50 held back for the thing you did not predict. Everything else can be free until the game is out and the numbers tell you what to buy next.
Keep the receipts in one folder from day one. When we went to set our price, having real spend figures turned an argument into arithmetic, which is also how we approached pricing the game on Steam.
FAQ
Do I need to pay the Steam fee to release a game?
Only for Steam. The Direct fee is 100 dollars per game and Valve returns it once that game earns 1,000 dollars in adjusted gross revenue. Releasing on itch.io costs nothing and lets you set the platform's cut yourself, which is why most first projects start there.
What is the cheapest way to get decent art?
Pay for the two or three images that sit on your store page and make everything else yourself in a consistent style. A simple style you can execute looks better than an ambitious style you cannot.
How much should a first game earn to be worth it?
Our first release earned under 400 dollars and I still count it as a win, because it taught us the whole pipeline from store page to patch. Judge the first one on what it teaches, and judge the second one on money.
Should I buy a better computer before starting?
Almost never. A five year old laptop runs Godot, Aseprite and Resolve. The one piece of hardware worth buying early is a cheap secondhand controller, because input bugs are invisible until someone plugs one in.
More in Studio
Studio
Handling Player Support as a Two Person Game Studio
Our second game sold about 1,200 copies in its first three weeks, and we were not ready for the mail that came with it.
Studio
A Game Release Checklist for a Small Indie Studio
Launch week is when small studios find out what they forgot.
Studio
How to Get Steam Wishlists Before Your Indie Game Launches
You put your game on Steam, shared the link once, and the wishlist counter has been stuck under 200 for a month.