Solynth Network — Essay
The games industry did not collapse. It moved. Understanding where it moved to is the difference between a bad four years and a good decade.
Start with the number that almost nobody reports, because it is the only one that reframes everything else. Between 2022 and 2026, according to data compiled by Amir Satvat through Always Supporting the Games Community, the global games workforce went from roughly 750,000 people to roughly 754,000. A net gain of 0.6%. Underneath that flat line, about 48,000 roles were eliminated and about 52,500 hires were made.
Fifty-two thousand five hundred hires. During what everyone correctly calls the worst labour crisis in the medium's history. The industry did not stop employing people. It stopped employing them in the same places, in the same structures, and on the same terms.
That is the whole argument of this piece, and everything that follows is detail. But the detail matters, because the strategic conclusions that fall out of it are not the ones the headlines suggest.
There is no honest optimism that skips this part, so here it is without softening.
The GDC 2026 State of the Game Industry, surveying more than 2,300 professionals, found 28% of respondents laid off within two years, rising to 33% in the United States, and half reporting layoffs at their current or most recent employer within twelve months. Two-thirds of AAA respondents saw layoffs at their company; a third of indie respondents did. Cumulative losses across 2022 to 2026 are approaching 60,000. Microsoft alone is cutting around 3,200 roles across its current fiscal year, with four studios divested and a fifth in unresolved consultation. Epic cut around 1,000. Ubisoft around 680, followed by a three-week strike at its Barcelona studio.
The competitive arithmetic is worse than the layoff arithmetic. Against those 52,500 hires stood roughly 288,000 people seeking games work over the same period — five seekers per hire globally, eleven per hire in North America. Twelve-month re-employment probability for someone with no prior industry experience runs under 5%. Typical searches run up to ten months, with more than a third of seekers still looking after a year.
Anyone telling you the conventional route is fine is either not looking at the data or is selling something. The conventional route — apply to studios, wait, apply again — is close to arithmetically hopeless as a primary strategy right now. Not because you are not good enough. Because there are eleven of you for every chair.
Here is where the pessimists stop reading and the strategists start.
If the odds of being selected are eleven to one, the rational response is not to compete harder for selection. It is to stop being a candidate and start being a counterparty. The industry has not run out of demand for games — quite the opposite. Steam generated around $17.7 billion in 2025, up 15% year on year, of which roughly $4.5 billion came from indie titles. What has run out is the willingness of large capital to carry permanent headcount through a development cycle.
Read that carefully, because it is the whole opportunity. The money is there. The payroll is what has gone. Studios still need games built; they have simply stopped wanting to employ the people who build them.
The structural consequence: a greater share of total production load is shifting to external development, co-development and contract work. Studios are relying on distributed talent more than at any point in the industry's history. That is not a job market. It is a supplier market — and suppliers are chosen on capability and relationship, not on getting through an applicant tracking system.
This is the single most important reframe available in 2026. In a labour market you are one of 288,000. In a supplier market you are one of however many outfits can actually deliver the thing, which is a vastly smaller number, and the barrier is demonstrated capability rather than a queue position.
Three places, none of them where the industry spent the last decade.
Geographically, east and south. Between 2024 and the first quarter of 2026, Asia-Pacific's share of job postings rose ten percentage points, with China driving most of it. Europe held flat. North America declined, its headcount down 11.5% across the four years, with California alone accounting for 47% of global games layoffs in 2025. The forecast trend is continued growth in Latin America, Eastern Europe and Asia. If you are working remotely and contractually rather than seeking a badge at a campus, this is a tailwind rather than a threat — the work is moving toward distributed delivery precisely because distributed delivery is cheaper than a Californian office.
Structurally, into contract and external development. Covered above, and worth repeating because it is the mechanism by which the other two operate.
Platform-wise, into user-generated content. This is the only genuinely new job category the crisis has produced. Roblox paid creators over $1 billion through its Developer Exchange programme in 2025, up from $923 million the year before. Fortnite creator payouts ran around $250 million. More significantly, the space is professionalising: Voldex acquired Brookhaven, Splitting Point acquired Grow a Garden, Gaijin bought The Pit on the Fortnite side. Where acquisitions happen, employment structures follow. A services layer has also emerged across India and Southeast Asia selling Roblox and Fortnite Creative development as contract work on the IT-outsourcing model.
Two cautions, because you should have them. Creator earnings follow a steep power law — Fortnite's average annual payout per creator is around $5,000, and an average that low with headline totals that high tells you exactly how concentrated the top is. And the services layer, while visibly real, has not been quantified by any credible source. Treat UGC as a genuine growth vector and a genuine lottery simultaneously.
Design is the hardest discipline in the industry to be hired into right now, and it is not close. Programmers have the easiest time; game designers face roughly nine times the difficulty, narrative designers and writers around twenty-two times. Satvat's explicit forecast going into 2026 was that the supply-demand gap in design, writing and narrative would get worse, with openings staying tight while demand rises.
That reads as a verdict against a design degree. It is not, and here is why.
The scarcity is in design positions, not design capability. Those are different markets. What the industry has surplus of is people who can be given a design brief and execute it — the employable, slottable, mid-pipeline designer. What it has almost none of is people who can define what should be built, why, and then cause it to exist. That has always been rare. It is rarer now because the studios that used to train that skill over fifteen years have stopped hiring the juniors who would have become those people.
Hiring has in fact concentrated into exactly two bands: early career and the fifteen-to-twenty-year experience range. The middle has been hollowed out. Which means the people who will occupy the senior design tier in a decade are, right now, mostly not inside studios at all. They are elsewhere, building. Whoever spends this period actually shipping — anywhere, on any platform, under any structure — arrives at the other side of the contraction with something the industry will by then be desperate for and will not have produced internally.
You need both halves. The first half is the opportunity and it is real. The second half is a live commercial risk and pretending otherwise would be doing you a disservice.
The opportunity: 36% of games professionals now use generative AI in their work, but the distribution is lopsided — 30% at game studios against 58% at publishing companies, support teams and marketing firms. The people closest to production are the slowest adopters. That gap is the opening. It means the tooling and workflow layer for actual game design is comparatively unoccupied, at exactly the moment when the industry's binding constraint has become the cost of carrying a production team. A capability that lets a small group produce at the scale that used to require thirty salaried people is aimed directly at the thing that broke.
The risk: 52% of games professionals believe generative AI is having a negative impact on the industry, up from 30% a year earlier and 18% the year before that. Only about 7% think it is positive, down from 13%. The hostility is sharpest in precisely the disciplines you would be selling into — 64% among visual and technical artists, 63% in game design and narrative, 59% among programmers. Students entering the field name AI-led displacement as a reason for their pessimism, and 82% of US respondents support unionisation, with the labour movement in games treating AI as an adversary.
What this means practically: the technical case will not be the hard part. The positioning will. A tool framed as replacing designers is walking into a market where two-thirds of the relevant profession has already decided against it, in a year when that profession is watching its colleagues lose their homes. The same tool framed as letting a designer who cannot get hired ship their own work anyway is aimed at the same people's actual problem, and is on their side of the argument rather than the other one.
That distinction is not spin. It is a genuine fork in what gets built, and it should be settled early, because it determines the product.
I would rather you go in knowing the sentiment data than discover it in a comment thread after launch. The hostility is not irrational and it is not going away on its own. It is navigable, but only deliberately.
Of everything above, two findings have effect sizes large enough to dominate the rest.
The first: getting even one to two years of industry experience raises the twelve-month re-employment rate from around 4% to around 30%. Not a marginal improvement — nearly an order of magnitude. Shipped, credited, real work is the single highest-leverage asset available, and it does not have to have been done inside a large studio to count.
The second: having a direct relationship with a recruiter or hiring manager correlates with hiring probability more than twenty times higher than having none. Satvat, to his credit, notes that this deepens the divide between the established and the not-yet-established. It does. It is also the most actionable finding in the dataset, and it argues that conference floors, Discord servers, union channels and the trade press are not adjacent to the work. They are the work, in a market this saturated.
The pessimistic reading of 2026 is that games is a shrinking industry. That reading is factually wrong: the workforce grew 0.6%, revenue is at record levels, and indie titles now make up the majority of the year's critically celebrated releases. What has actually happened is that the industry has stopped being a place where you are given a career and become a place where you have to construct one — earlier, more deliberately, and with more ownership than any previous generation of developers needed.
That is genuinely harder. It is also, for the specific set of people who can do it, better. Ownership beats employment on every axis except security, and security is the one thing the studios have stopped offering anyway. The trade is no longer available. Given that, the question is not whether to build rather than apply. It is only how fast.
You have the design training, the tooling, and a platform to put it on that does not require anyone's permission. That is a materially stronger position than 288,000 other people are in this morning, and the reason it is stronger has nothing to do with luck.