48% of game studio founders are currently operating at a loss.

That's from the tenth annual games industry employment survey by Values Value and InGame Job. Alongside it, 60% of founders say keeping the team small is a deliberate strategy, and 35% have skipped a hire because AI could do the work.

I understand those decisions. What concerns me is what gets cut when a team stays small, because its almost always the same thing. Marketing, community and publishing produce nothing you can play, so they read as overhead and go unfunded or come as a last thought.

Roughly 12,000 games launched on Steam in the first half of 2026, up 19% year over year. And Steam is not in the business of guaranteeing you new player discoverability.

"Hire a marketing team" is not advice, it's a budget many of these studios simply cannot afford. To me there are four real options:

1. Build an internal team. Full control and people who know the game intimately. The cost is salary plus benefits plus taxes. Then what do those people do after launch when the workload drops and so does the revenue.

2. Sign with a publisher. Development funding, a go-to-market budget and the people to spend execute it, which is a strong package if you need all three. You give up revenue share, often with the publisher recouping first, so you can ship a commercial success and wait a long time for dollar one. You also become one title in a portfolio, you do not set the budget, and guaranteed marketing spend is almost never written into the contract.

3. Hire a traditional agency. Excellent at project work, so if you need a trailer or a paid campaign, this is the right call. What you do not get is a foundational partner. They are not in your Discord on a Tuesday, not answering replies to your posts, and not in the room for the hard strategy calls.

4. Fractional embedded support. Senior marketing and publishing people working inside your team, on your roadmap, at a monthly cost rather than a headcount cost, covering both the always-on work and the campaign moments. You keep the revenue and the decisions.

Ill admit I’m not neutral, since that last model is what we built Ayzenberg Accelerate to do. I think this is a funding question first and a control question second. A team that needs a development check should be talking to publishers or other funders. A funded team has options and many times a publisher is not the right one.

I do know this, deciding that go-to-market is what you figure out in the last ninety days rarely works out the way you want it to.