Opinion.
Every few years, a company backed by billions in venture capital announces it will dethrone Steam. Every few years, it fails. Google Stadia is dead. Amazon Luna is on life support. The Epic Games Store has spent over $1 billion in exclusivity deals and free game giveaways and still holds less than 10% of the PC gaming market. Meanwhile, Valve’s platform just set a record of 42 million concurrent users in January 2026 and controls roughly 75% of PC digital distribution globally. The pattern is not a coincidence. It is a structural consequence of who owns what, and why platform enshittificationThe gradual degradation of online platforms as they shift loyalty from users to advertisers: first attracting users with a good product, then extracting value from them, then abandoning them entirely in favor of revenue. spares the companies that refuse to play the growth game.
The Platform Enshittification Cycle, Briefly
Cory Doctorow coined the term “enshittification” in 2022 to describe the lifecycle of investor-backed platforms. The pattern is predictable. First, a platform subsidizes users to build a base: free games, below-cost hardware, generous revenue shares for developers. This is the “be good to users” phase, funded by investor cash. Second, once the user base is locked in, the platform shifts value from users to business customers (advertisers, publishers, data buyers) to demonstrate revenue growth. Third, the platform extracts from everyone, including its own business partners, to satisfy shareholders demanding returns. Doctorow’s 2025 book expanded this into a unified theory of why platforms degrade systematically rather than accidentally. The mechanism is not greed in the moral sense. It is the rational response to a specific ownership structure.
The Graveyard of Steam Competitors
Google launched Stadia in 2019 with the full weight of Alphabet’s resources behind it. The technology worked. The business model did not. Google needed Stadia to justify itself to shareholders within the same quarterly cadence that governs Search and YouTube ad revenue. When user adoption was slower than projected, Google did what publicly traded companies do: it cut. The internal game studios were shuttered in 2021. The entire service was killed in January 2023, barely three years after launch. Google refunded all purchases, which was decent of them, and also an admission that the thing never had enough momentum to be worth keeping alive.
Amazon Luna followed a similar trajectory. Backed by a company whose every division must eventually contribute to the bottom line that Wall Street evaluates each quarter, Luna launched with a fragmented subscription model that confused users. By 2023, it was hemorrhaging titles. The service still technically exists, in the way that a restaurant with no customers still technically serves food.
The Epic Games Store is the most instructive case because Tim Sweeney, to his credit, has tried to resist the pressure. Epic remains private, but not independent: Tencent holds a 40% stake acquired in 2012. Sweeney maintains voting control and has publicly stated that Tencent does not direct creative decisions. But Epic’s valuation dropped from $31.5 billion to $22.5 billion between 2022 and 2024. That kind of decline, even in a private company, creates pressure. Investors expect returns. The Epic Games Store’s strategy of buying exclusives and giving away free games is a classic phase-one platform enshittification move: subsidize the user to build the base. The question is what comes after. Epic projects $6 billion in revenue for 2025, largely from Fortnite and Unreal Engine licensing, not from the store itself. When the subsidy engine runs dry, the store must either become profitable on its own terms or start extracting more from the developers and users it courted.
Why Valve Is Structurally Immune
Valve Corporation has never taken external investment. It has never held an IPO. Gabe Newell, co-founder and president, holds a controlling majority stake estimated at over 50%, with the remainder distributed among long-term employees through an internal equity program. There are no quarterly earnings calls. There are no institutional shareholders demanding 15% year-over-year growth. There is no board of directors answering to pension funds and hedge funds.
This is not a minor organizational detail. It is the single most important fact about why Steam operates the way it does. Platform enshittification requires a specific trigger: the point at which the platform’s owners need to extract more value than the platform naturally generates. For publicly traded companies, that trigger is constant. The stock price must go up. Revenue must grow. If organic growth slows, you manufacture it by squeezing users or business partners. For Valve, that trigger does not exist. The company generated an estimated $16.2 billion in revenue through November 2025. It is phenomenally profitable. And it answers to nobody except its own employees and a founder who has repeatedly stated he prefers it that way.
The results are visible in decisions that no investor-backed company would make. Steam’s refund policy lets users return any game played for less than two hours within 14 days, no questions asked. The Steam Deck, which has sold roughly 4 million units, ships with full user access to the hardware: you can install Windows, swap the SSD, run competing stores on it. Valve does not lock you into its ecosystem because it does not need to. When your business model is “be the best option and take a cut,” you do not need to trap anyone.
The 30% Question
Critics, particularly Epic, point to Steam’s 30% revenue share as evidence that Valve is extracting rent from its dominant position. Epic takes 12%. The argument seems straightforward: Valve charges more because it can.
But platform enshittification is not about the size of one fee. It is about the direction of change over time. Steam’s 30% cut has been its standard rate since 2003. In 2018, Valve introduced tiered rates: 25% after $10 million in sales, 20% after $50 million. The fee has only moved in one direction, and that direction is down. Compare this to the typical investor-backed trajectory, where fees start low to attract users and then climb once the platform achieves dominance. Apple’s App Store launched at 30% and has since added a 15% tier for small developers, but only after antitrust pressure and the threat of legislation, not out of structural generosity.
The difference is not that Valve is morally superior. It is that Valve’s ownership structure does not require the fee to increase. A 30% cut on $16 billion in annual revenue, with no investors demanding a return, is a fundamentally different animal than a 30% cut that must grow to justify a stock price. One is stable. The other is the starting position of a ratchet.
The SteelmanA rhetorical technique where you present the strongest possible version of an opponent's argument before refuting it. The opposite of a straw man.: Maybe Investors Bring Discipline
There is a genuine counterargument here, and it deserves honest treatment. Private companies with no external accountability can become complacent. Valve notoriously ships products on “Valve Time,” with delays measured in years. Half-Life 3 became a meme precisely because no shareholder could force the company to finish it. The flat organizational structure that Valve champions has also been criticized by former employees as chaotic, with projects starting and dying based on internal politics rather than strategic planning.
Investor pressure, when it works well, provides discipline. Deadlines. Resource allocation. The obligation to ship. There is a reason most of the world’s most productive companies have external shareholders: accountability structures can be genuinely useful.
But the question is not whether investor pressure has benefits. The question is whether those benefits outweigh the platform enshittification pressure that comes bundled with them. For platforms specifically, the evidence is not encouraging. The discipline that investor pressure provides tends to optimize for growth metrics (monthly active users, average revenue per user, engagement time) rather than for the thing that makes a platform worth using in the first place: being good for the people on it. When the metric becomes the target, it ceases to be a good measure. And when the metric is quarterly earnings, the platform’s users become the resource being mined.
Why This Compounds
The structural advantage of private ownership does not just persist. It compounds. Every year that Steam avoids platform enshittification, its library of user-owned games grows. Every user who has 500 games on Steam faces a switching costThe cost or friction a user faces when moving from one platform to another, including time, money, and effort invested in the original. Also called switching barriers. that no competitor can overcome with free giveaways. This is sometimes described as lock-in, but it is a peculiar kind: Valve did not engineer it through DRMDigital Rights Management; technology that restricts how users can copy, share, or modify protected digital content like software, music, or video. or proprietary formats. It emerged naturally from being a stable, functional platform for two decades.
Meanwhile, every year that an investor-backed competitor fails to dethrone Steam, the pressure to show returns intensifies. The Epic Games Store’s free game strategy costs money. Tencent’s 40% stake is not a charitable donation. At some point, the math demands a change in strategy, and that change will not favor users. This is not speculation. It is the documented pattern of every investor-backed platform that has reached maturity. The only variable is timing.
Steam’s competitors are not losing because their products are worse. Some of them offered genuinely good technology (Stadia’s streaming was impressive) or better financial terms for developers (Epic’s 12% cut). They are losing because their ownership structures create an inevitable conflict between what users want and what investors need. Platform enshittification is not a choice individual executives make. It is the predictable outcome of a specific set of incentives. Valve avoids it for the same reason a privately held family restaurant does not become a fast-food franchise: nobody is making them.
The lesson is not that private companies are inherently virtuous. It is that the war against Steam was always unwinnable for the specific entities that tried to fight it, because the weapons they brought, investor capital, created the very vulnerability that makes them unable to compete long-term. You cannot out-user-focus a company that has no one to answer to except its users. The structure is the strategy. Everything else is a subsidy with an expiration date. And as with any industry built on extracting value from its audience, the extraction eventually becomes the product.
Sources
- Enshittification (Wikipedia overview, with primary sources in footnotes)
- Steam Statistics 2026: Market Share, MAU and Revenue (DemandSage)
- Valve Corporation: Ownership and Corporate Structure (Wikipedia)
- Epic Games: Ownership, Tencent Stake, and Financial History (Wikipedia)
- Analysis: The Impact of Google Stadia Shutdown on Cloud Gaming (GeekWire)
- The Cory Doctorow Doctrine: Enshittification (Washington Monthly, 2025)
- Steam Deck Sales Estimates: 4 Million Units (IDC via Famiboards)
