When you pressure a competitor into solving a problem they never planned to solve, you might just disrupt yourself. The Valve–Proton story is Christensen’s Innovator’s Dilemma in real time — told by someone who saw it coming from the inside.
After a long break, I’ve been spending more time with Linux again lately — and two things brought me back. The first is the steady erosion of privacy and data security at the big platform providers: what used to feel like acceptable trade-offs increasingly doesn’t. The second is more mundane but more immediate: Apple has announced end-of-life support for Intel-based MacBooks by the end of this year. My old machine — perfectly capable hardware — is being pushed out of the ecosystem not because it can’t run the software, but because that’s what the roadmap requires. Sound familiar? Windows 10 users are living through the exact same story right now, except Microsoft is even more aggressive about it: millions of machines that run fine are being declared obsolete to push users toward a hardware refresh cycle driven by software-strategic interests, not technical necessity.
Linux gave my old MacBook a second life. But that’s a post for another day. The point here is broader: for everyday office users, moving to Linux has become a surprisingly manageable step. LibreOffice, browser-based tools, cloud storage — the friction is lower than it’s ever been. Gamers, however, have long faced a harder problem. Their libraries, their saves, their anti-cheat dependencies — all of it tied to Windows. That’s the lock-in that made switching feel impossible.
What changed that — and why it happened the way it did — is one of the more interesting product strategy stories of the last decade. It starts with a Microsoft decision that backfired, a former Microsoft employee who saw it coming, and a compatibility layer that quietly made Windows optional.
In 2012, Microsoft launched the Windows Store with Windows 8. It was, by most accounts, a commercial disappointment. But it sent a message that Valve — the company behind Steam, the dominant PC gaming platform — took very seriously: Microsoft might one day lock third-party storefronts out of Windows entirely.
Gabe Newell, Valve’s co-founder and CEO, said so publicly at the Casual Connect conference in Seattle that July. And his alarm wasn’t the reaction of an outsider reading the tea leaves. Newell had spent nine years inside Microsoft before co-founding Valve in 1996 — including formative years working on the early versions of Windows itself.
Who is Gabe Newell?
Newell joined Microsoft in 1983 straight out of Harvard and worked there for nine years, contributing to early Windows development. He left in 1996 to co-found Valve, which released Half-Life (1998) — one of the most influential games ever made — and later built Steam into the world’s largest PC game distribution platform with over 130 million active users. When he called Windows 8 a catastrophe, he wasn’t speculating. He was pattern-matching against a company culture he knew intimately.
That insider context matters enormously. Newell understood Microsoft’s strategic logic not from press releases but from having lived inside the incentive structures, the product thinking, the platform ambitions. When he looked at Windows 8 and the Windows Store, he wasn’t seeing a failed UI experiment. He was seeing the opening move of a platform control strategy — the same kind of thinking he had watched Microsoft apply to other markets for a decade.
Most competitors react to what a company does. Newell could anticipate what Microsoft would logically do next — because he understood not just the product but the strategic DNA behind it. This is why his Linux pivot happened in 2012, years before the threat fully materialized. Ars Technica noted at the time that his “negativity for Windows 8 and the future of Microsoft was stunning” to those who visited Valve’s campus that spring. He wasn’t reacting. He was already three moves ahead.
“I think Windows 8 is a catastrophe for everyone in the PC space. We want to make it as easy as possible for the 2,500 games on Steam to run on Linux as well. It’s a hedging strategy.”
— Gabe Newell, Casual Connect, Seattle, July 2012
And rather than lobby, complain, or hope for the best, Valve did something far more consequential: they started treating Linux as a strategic hedge. The result was Proton — a compatibility layer built into Steam that lets Windows games run on Linux without modification. And in building it, Valve didn’t just protect themselves. They quietly dismantled one of Microsoft’s last meaningful consumer lock-ins.
The lock-in nobody talked about
To understand why this matters, look at the numbers.
Back when Newell gave that speech, he was already targeting 2,500 Steam games for Linux ports. That was the scale of what he was planning.
For years, the unspoken reason millions of people stayed on Windows wasn’t Office. It wasn’t Active Directory. It was their Steam library. Enterprise software had been moving to the browser for over a decade. Google Workspace was eating into Microsoft 365’s market share. File management, communication, even development — all increasingly OS-agnostic. But gaming was different. Games are native applications. They depend on DirectX, Windows APIs, anti-cheat software tied to the kernel. Your 400-game Steam library? Windows only.
That was real lock-in. Quiet, powerful, and almost never discussed in analyst reports. Microsoft, apparently, didn’t discuss it much either.
The Apple blueprint — what platform control actually looks like
To understand what Microsoft was trying to build — and why Newell, with his insider knowledge, was right to be alarmed — it helps to look at what Apple had already pulled off. iTunes, launched in 2003, is the clearest precedent for what happens when a company controls both the OS and the gateway to applications and content.
Within three years of launch, iTunes had captured roughly 80% of the U.S. digital music download market. The record labels, desperate to counter piracy after CD sales dropped for the first time ever in 2000, had handed Jobs a complete monopoly on the terms — 99 cents per track, Apple’s pricing, Apple’s DRM, Apple’s rules. As one analyst put it at the time: “Can the record labels really afford to pull their catalog from iTunes?” They couldn’t. Apple collected 30% of all revenue, dictated pricing, controlled distribution, and locked music to Apple devices via FairPlay DRM. The labels had traded one set of problems (piracy) for another (permanent dependence on a single gatekeeper).
The App Store, launched in 2008, applied the same logic to software. Apple’s 30% commission on all digital sales became the default toll on the entire mobile economy. Epic Games eventually challenged this in court — and the fight revealed the structural reality: whoever controls the platform gateway controls the business model of everyone who builds on top of it.
“Apple has locked down and crippled the ecosystem by inventing an absolute monopoly on the distribution of software, on the monetization of software.”
— Tim Sweeney, CEO Epic Games, 2020
This is precisely what Newell — who had helped build the very OS Microsoft was now trying to weaponize as a platform — saw coming for PC gaming in 2012. A Windows Store with Xbox Live integration, controlling which games could be distributed on Windows, taking a cut, setting the terms. Game studios would either comply or lose access to the largest PC gaming market on earth. Newell had seen this playbook from the inside. He knew it was real. And he acted accordingly.
Enter Christensen — and the Nokia lesson Microsoft should have learned
Clayton Christensen — Harvard professor, author of The Innovator’s Dilemma — spent his career documenting how well-managed companies fail. Not through incompetence, but through competence. They listen to their best customers, optimize their strongest products, and completely miss the smaller player entering from an adjacent market.
The pattern is consistent: the incumbent doesn’t fail because they made the wrong decision. They fail because they made the rational decision — and the attacker changed the game.
The Nokia story is the textbook example — and Microsoft is more deeply implicated in it than most people remember. Nokia was the dominant handset manufacturer when the iPhone launched in 2007. Their response was dismissive: no physical keyboard, not a real business device. They doubled down on Symbian and their existing platform logic. What they missed wasn’t the iPhone itself — it was the shift from hardware-defined value to platform-defined value.
What makes the story even sharper is the acquisition Microsoft didn’t make. At the time, BlackBerry had exactly what Microsoft needed to compete in mobile: an established enterprise user base, a reputation for security, deep integration with corporate IT infrastructure, and push email that entire industries had built workflows around. A Microsoft acquisition of BlackBerry would have combined BlackBerry’s proven enterprise ecosystem with Microsoft’s Exchange, Active Directory, and Office dominance — a genuine platform play built on real, compounding synergies. Instead, Microsoft acquired Nokia in 2013 for $7.2 billion — a company whose ecosystem had already collapsed, whose OS was going nowhere, and whose brand was in freefall. They bought a corpse and tried to reanimate it with Windows Phone.
The strategy failed comprehensively. By 2016, Microsoft had written off nearly the entire acquisition, laid off most of the Nokia workforce it had absorbed, and quietly exited the consumer smartphone business. Microsoft didn’t just observe Nokia’s strategic error from the sidelines — they bought into it, doubled down on the same flawed logic, and lost billions in the process. Meanwhile, they had passed on BlackBerry — a company whose enterprise ecosystem could have been genuinely potentiated by Microsoft’s existing corporate infrastructure, where the M&A synergies were real and compounding rather than speculative.
Both the acquisition they made and the one they didn’t tell the same story: Microsoft consistently evaluated mobile through the lens of its existing business model, missing what the market had actually become. And yet, simultaneously, they were trying to apply that same platform-control logic to PC gaming via the Windows Store — without recognizing that a former insider was watching, and already building an escape route.
Microsoft saw Valve as a game distributor. Not a threat to Windows’ OS hegemony. So when they pushed toward a more controlled app ecosystem, they weren’t thinking about what a former Microsoft insider — who understood their strategy better than most external analysts — might do in response. Microsoft abandoned its Windows Store gaming strategy entirely in 2019, having failed to gain traction. By then, Proton was already in motion.
The technical play that changed everything
Proton is built on Wine, a long-standing compatibility layer for running Windows applications on Linux. Valve supercharged it with DXVK and VKD3D-Proton, which translate DirectX calls — Microsoft’s proprietary graphics API — into Vulkan, an open standard that runs natively on Linux.
Here’s the irony: Vulkan was designed from scratch, without decades of legacy constraints. It’s leaner, closer to the metal, and in many benchmarks outperforms DirectX on equivalent hardware. Microsoft’s sprawling API, carrying years of backward-compatibility overhead, is being outrun by an open standard built on a clean slate. The compatibility layer that was supposed to be the weak point turned out to be negligible — most Steam games now run on Linux within 0–5% of their Windows performance. Sometimes faster.
The Steam Deck, launched in 2022, made this mainstream. Millions of users are now running Linux-based gaming systems without knowing it. Developers who want their games Verified for Steam Deck must ensure Linux compatibility. The flywheel turned.
What this means for Microsoft now
Windows still holds around 80% of the PC gaming market — but that number was effectively 95%+ in 2012. Linux gaming has grown from a rounding error to a structurally significant segment. More importantly, it has removed the psychological barrier. Switching is now possible. And meanwhile, Microsoft’s growth bets — Azure, Copilot, Microsoft 365 — all face brutal competition without the monopoly margins that Windows once provided.
The company that bought Nokia to win mobile, launched a Windows Store to win app distribution, and pushed Windows 11 hardware requirements to accelerate device refresh cycles is still playing the same game. The question is whether the lesson from all three — that platform control attempted too aggressively creates the very alternatives it seeks to prevent — has been internalized. The evidence so far suggests not.
PM takeaways
Think in second-order consequences. Microsoft’s Windows Store strategy was defensible in isolation. The failure was not modeling what a pressured competitor — especially one with insider knowledge — would do in response.
Never underestimate a founder who worked at you. Newell didn’t guess at Microsoft’s intentions. He understood the strategic logic from the inside. When a key competitor has deep institutional knowledge of how you think, their threat model of you is more accurate than yours of them.
M&A synergies are only real if the ecosystem is alive. BlackBerry had compounding synergies with Microsoft’s enterprise stack. Nokia had a collapsing one. Buying the wrong asset at the wrong moment doesn’t just fail — it crowds out the right move.
Platform control attempted too aggressively creates the alternatives it seeks to prevent. Nokia’s failure didn’t stop Microsoft from doubling down on the same platform logic in mobile — and losing $7.2 billion. Valve saw the same logic being applied to PC gaming and built the escape route. The pattern is consistent.
Constraint drives innovation. Proton exists because Valve had no choice. Some of the most durable product investments come from survival necessity. Know which of your competitors are currently being backed into a corner — and what they might build to escape.
Lock-in is fragile when it’s invisible. Gaming was Windows’ most powerful consumer lock-in precisely because nobody talked about it — including Microsoft. Your most important moats are often the ones your own team doesn’t recognize as moats.
Read Christensen. The Innovator’s Dilemma is 30 years old and still describes what’s happening in this industry every quarter. The pattern doesn’t change. The players do.
Microsoft didn’t lose gaming overnight. Game Pass, Xbox, and first-party titles are real strategic assets. But the silent lock-in that kept hundreds of millions of users on Windows because their game library demanded it? Valve broke that. Because Microsoft left them no choice. And the person who saw it coming had spent nine years learning how Microsoft thinks — from the inside.
Sources: Clayton Christensen, The Innovator’s Dilemma (1997). Gabe Newell quote sourced from Casual Connect, Seattle, July 2012, as reported by Engadget and Ars Technica. Windows gaming market share data via Mordor Intelligence (2025). iTunes market share via NBC News / Nielsen SoundScan. Epic v. Apple quotes via Wikipedia / CNBC. ProtonDB (protondb.com) tracks Linux game compatibility in real time.
Aye Stephen is an accomplished Product Leader and has been last working as Chief Product Officer (CPO) at one of Europe’s leading private equity backed eCommerce ERP solutions. With a strong background in product and project management and leadership coming from 20 years experience, he is an expert in building high performing product teams in agile environments and organizational change management. Stephen holds an MBA from Goethe Business School Frankfurt and an M.A. in American Studies/Media Science from Philipps University Marburg.