Earlier this summer, I penned a piece for Dean Takahashi at GamesBeat about how Xbox was Microsoft’s new burning platform. A friend read it and helpfully suggested the next one should include “the answers.” Right you are, sir.
Microsoft has given me a lot more to work with since then. In July, Xbox announced roughly 3,200 job eliminations across fiscal 2027 and plans to move four studios out of the company. On September 22, it eliminated a further 268 roles under that plan and said the restructuring, including the studio divestitures, was roughly three-quarters complete. More importantly, Microsoft redrew the studio map: Activision will oversee Rare and World’s Edge and build the next Halo; Bethesda gets Obsidian; King gets Microsoft Casual Games; and Playground and Turn 10 will combine into one studio.
Before I dive into what that means, let me say this: behind every line in that org chart are capable people who did what they were asked to do. Their human cost is worth acknowledging first and foremost.
Xbox CEO Asha Sharma wrote that her number one priority was “CORE: Strengthen our platform, led by console.” New chief strategy officer Matthew Ball has been even more forthright: console number one is step one of stabilizing the console business.
Which makes sense, but leads to the question, “If and when the console is stabilized, then who and what is Xbox for?”
Before I answer that, a couple of distinctions are required. Microsoft Gaming is the entire business and associated P&L: King, Activision, Blizzard, Bethesda, Mojang, hardware, subscriptions, stores, and services. The Xbox platform is the relationship with players and developers: the console, Xbox Mode, commerce, identity, ownership, community, tools, and third-party support. I’ll use XboxOS as shorthand for the OS component of that platform; Microsoft has announced no such product. Microsoft Gaming may grow while the Xbox platform withers.
Quo vadis, Xbox? Whither goest thou?
Part I — From console anchor to XboxOS
A prophecy in search of a prophet
Way back in 2015, after repairing the mess that was the Xbox One launch, Phil Spencer told Polygon that Xbox had to “move beyond fixing things.” They needed a vision that gamers could buy into, teams could execute, and partners could support. He was also careful to draw the distinction between putting gamers at the center versus putting “a piece of plastic or a specific device” at the center.
Eleven years later, most of that vision’s reach is real. Xbox publishes on PC. Play Anywhere, cloud, and cross-play exist. Phil’s prophecy was fulfilled in fragments, but the fragments weren’t a compelling reason to choose Xbox. “Play anywhere” tells me where the game can be. It doesn’t tell me why I should choose Xbox to play it on.
The internal logic is simple. Windows was always about ubiquity. Azure was about cloud usage. The console business was about stores. Game Pass was about subscriptions. Studios were about hits. Each team could (and did) make a perfectly rational decision, but taken together the whole enterprise was incoherent. Meanwhile, Steam had one focus: the PC customer. Xbox never answered the two questions that mattered: which layers must we own, and which internal business would take priority (and when)? No executive had both the authority and the incentive to break that tie. And so Xbox became everything, and nothing; to everyone and nobody.
Asha’s reset memo reads like a confession. Silos. Independent functions. Decisions that went through as many as 14 management layers. The “platform team” grew by 40% while players and playtime fell. Finally, “for the first time,” a single executive has end-to-end P&L responsibility for content, hardware, platform, and services.
I can see the logic in the new organization. Fewer business units. Put franchises under operators with proven commercial machinery. It’s a plausible answer to an execution and margin problem. But it is still a publishing answer, not a platform answer. Nothing in this reorganization explains how those units will create gravity for the Xbox platform rather than optimize their own audiences and P&Ls.
Enough history. I go on at length here because the same managerial conflict that derailed the last strategy can just as easily derail the next. Managerial and divisional conflict is baked into Microsoft’s DNA.
Console comes first
Xbox can make hardware. It can publish games. It can sell subscriptions. It can run mobile businesses. It can use the cloud. It can run a platform. What it cannot do is make all six the organizing principle.
Console is the anchor. Asha says it drives the most revenue, is the flagship experience, and is at the center of Xbox fandom. The numbers explain the need to emphasize this: content and services revenue was down 10% for the June quarter, and full-year hardware revenue was down 29% due to lower console volume.
CNBC noted that Xbox was expected to run at an internal margin of about 3%. Microsoft provided the more illuminating comparison: Xbox runs three to 10 times below comparable platform and publishing businesses.
The doctor’s immediate Rx is simple: stop the bleeding. Give players a reason to buy an Xbox and reassure customers who wonder where their libraries and communities will end up on a platform in managed decline.
My friend Tadhg Kelly has offered the clearest formulation of the console-first case yet. Microsoft won’t dislodge Steam from its network effects, trusted library, community, discovery engine, and decades of habit. Owning Windows isn’t the same thing as owning PC gaming. His advice is admirably surgical: build the best console, and the best games.
True, but implied in this is that Xbox must ship a signature product. Without the anchor, the Xbox “platform” is simply a suite of services customers can purchase piecemeal from elsewhere.
I was there at the beginning of Xbox. Inside Microsoft, the fear wasn’t merely that Sony would sell more consoles; it was that another company would gain ownership of the operating and commerce layer in the living room. First-party games were how Xbox earned the right to exist. They were not the goal of the exercise.
Brand works the same way. Players build permanent libraries, communities, subscriptions, and habits where they expect the platform to exist. If Xbox hardware is a short-term play, if exclusivity reverses from year to year, or if today’s stores are irrelevant to tomorrow’s Microsoft, customers will go elsewhere; stores are now a commodity. And developers will have no choice but to follow their players as they exit.
Signature games matter even when the P&L on an individual title is less thrilling. They tell customers Microsoft has their back. OG Xbox didn’t need to own every studio; it needed enough signature games plus adequate third-party support to create “gravity”.
This year’s Gamescom featured a plausible split: Gears of War: E-Day and Fable were demonstrated for Xbox console, PC, and Game Pass with no PS5 version announced on either platform. State of Decay 3 will also be on PS5, with Minecraft Dungeons II remaining fully cross-platform. Microsoft also agreed to publish Kojima Productions’ PHYSINT after PlayStation Studios walked away from it. That’s the kind of outside bet Xbox should be making: a potentially defining game without the need for another studio acquisition. The platforms weren’t announced, so it proves nothing about exclusivity.
The broader point is consistency. Xbox can publish some games everywhere. They need enough games that define their own platform, with a policy customers can believe for longer than one news cycle.
Halo is now the clearest test. A purpose-built Activision team may make a better Halo (franchise stewardship isn’t sacred to an org chart). But Halo has never been just another IP cash flow. It was the flagship. It helped Xbox earn the right to exist. If the new team is measured primarily on reach and franchise revenue, maximum distribution is the rational answer. If Xbox needs Halo to create platform gravity, somebody with authority over the whole P&L must be willing to value that contribution. Same franchise. Different objective.
A better console is necessary to stabilize the anchor. But, as with the Xbox at its inception, the goal needs to be bigger.
RAM-a-geddon and the hardware trap
Tragically, this may be the worst possible moment in console history to make the box the linchpin of the strategy.
Microsoft’s latest console price hikes went into effect August 1st. The disc drive 1TB Xbox Series X is now $799.99 in the United States. The 512GB Series S is $499.99 and the 1TB is $599.99. Microsoft stated that storage and memory costs have increased over 2.5 times and they expect another doubling by fall 2027. Which brings them to the awkward, undeniable reality that unlike phones or PCs, consoles are usually sold for less than they cost to make.
Gamescom provided a glimpse into the problem. Microsoft opened pre-orders for the 1TB Xbox Series X25 collector’s edition console for $899.99 with Halo: Campaign Evolved included. Now, a collector’s edition console isn’t a mass-market pricing strategy, but $900 is still a daunting number for a console. Sharma told the BBC that Helix has to balance “efficiency and affordability alongside performance.” She reiterated commitment to backward compatibility and player choice… and then failed to confirm a disc drive would be a part of the next system.
Steam Machine from Valve offers another side of this grim equation. It starts at $1,049 for 512GB and goes up to $1,349 for 2TB. One of Valve’s engineers said in July that bulk component pricing was “still getting worse” and that retail pricing lags the supply market by three to six months.
RAM-a-geddon or not, every hardware maker is living with this costly new reality.
The Helix Project
The Helix Project aims to balance three competing imperatives: maintaining enough Xbox Store business to subsidize the console, opening the platform enough for the existence of cross-platform play to mean something, and keeping it affordable enough to be a mass-market console. Microsoft has Game Pass, first-party sales, and accessories to help pay the bills. But none of them change the fundamental subsidy vs openness calculus.
Microsoft-built boxes approaching $1,000 are poor foot soldiers for winning back players, much less attracting new ones. The decisive contest will be XboxOS versus SteamOS, and hardware is only one front in that fight.
The Xbox console must remain the reference device and brand anchor: the best integrated version of the Xbox experience (for those who can afford it). XboxOS and the services around it are how that experience gets to be something other than a box Microsoft makes. “Console first” does not mean “console only.” It also does not mean that calling the operating system the product makes hardware irrelevant. It means that the economics of the platform can’t depend solely on Microsoft making every box itself.
“This is an Xbox” was a campaign worth aiming for. The problem was timing and execution. People must believe the promise you’re making, and the experience must deliver. “This is an Xbox” did neither. And so it became the punchline to a joke Microsoft told on itself.
Why platform over pure publishing?
Consoles are expensive and risky. And Microsoft already has King, Call of Duty, and Minecraft. Why not give up on platform ambition and just become an excellent multiplatform publisher?
Because owning the platform is more valuable than renting space on everybody else’s.
Microsoft, of all companies, should know this.
Owning the platform means owning the customer: identity, commerce, entitlements, discovery, and transaction data. It means setting developer terms and gaining leverage over Valve, Sony, Apple, and Google. It means keeping the platform margin. In blunt economic terms, the platform owner is the price-maker.
The pure publisher is the price-taker. Apple, Google, Sony, and Valve determine the terms and the placement, and they know more about your customers than you do.
Microsoft would still have excellent content, it just would be renting every important storefront and having to pay a toll at every gate.
The platform prize is practical for Microsoft. Cross-device identity and ownership become much easier when Microsoft owns the entitlement layer. Microsoft can create one clear path for developers to build, publish, and manage rights across Xbox console and PC. PlayFab, Azure, Game Pass, first-party commerce, and the player’s library can all reinforce one another instead of behaving like separate fiefdoms.
There’s a hard stop to all this. Microsoft has to own a scarce layer that players, developers, and hardware makers all willingly value. If the PC Store never competes, if Play Anywhere remains a niche offering, and if XboxOS offers nothing that Steam and regular old Windows doesn’t already, then the platform case is dead. In which case rationally pruning the expensive pieces and harvesting the cash from King, Minecraft, and Call of Duty is perfectly sensible. A platform can’t be memo’d into existence; it has to earn (and keep earning) that right.
The operating system is the product
Valve provides the useful precedent here. They don’t make every device that runs SteamOS. Lenovo released the first licensed third-party SteamOS handheld device in May 2025. Valve can have hardware makers build the boxes while Valve maintains the store, identity, library, and transactions at the center.
Microsoft is slowly following that path. Xbox Mode started rolling out to Windows 11 PCs in April. Project Helix is designed to play both Xbox console and PC games. Play Anywhere has more than 1,500 titles. Microsoft says developer onboarding that previously could take months can now be done in about 30 minutes. Publishing is still a separate step, but the GDK and documentation are public.
Sharma’s choice of words at Gamescom was telling. She referred to Helix as a “great family of devices,” not simply the next console. It could be a Microsoft-built reference machine, hardware from OEMs, or both; Sharma didn’t say. But the language was nonetheless a step in the right direction. Microsoft defines the experience, identity, entitlements, commerce, and certification while partners can focus on building the hardware. No OEM or licensing terms have been announced, so this was a signal rather than a plan reveal.
Gamescom also offered the first concrete test of the ownership promise with Disc to Digital. A player will be able to use a supported Xbox One or Series X console disc to claim a digital entitlement, including Play Anywhere and cloud benefits when the game supports them. Xbox Insiders started testing on August 31st. The catalog, publisher participation, transfer rules, and resale treatment will all need to be proved over time.
Nonetheless, Microsoft is finally treating ownership as a platform asset. If this works, a supported disc becomes an entry point into the entitlement system rather than a dead end attached to one drive. Physical ownership could be preserved through the Helix era, regardless of whether the drive is internal, optional, external, or even absent. Microsoft hasn’t said which. Sony’s decision to go discless for new PlayStation releases in 2028 provides Xbox with a conspicuous hole to wedge through. Combined with the PHYSINT agreement, it’s an inexpensively player-friendly alternative to a rationally retreating competitor. The opportunity matters only if Microsoft turns it into a durable policy on content and ownership.
These are real improvements. Xbox Mode is still on top of Windows with all its seams. In Ars Technica’s tests, purchasing or downloading from Steam or Epic usually meant dropping back into their desktop apps. Microsoft considers this flexibility. With a controller in my hand on the couch, it feels rough and unfinished.
Now, put yourself in Lenovo’s or Asus’s shoes. Why would either company choose XboxOS over plain Windows or SteamOS?
Microsoft would need to offer a compelling bundle to get OEMs to switch: a seamless controller-first user interface, the Xbox console and PC catalogs, Play Anywhere cross-buy and cross-save, Game Pass, compelling cross-certification, less integration than regular Windows, co-marketing, and licensing terms that leave room for the OEM to make money. That’s a tall order.
Can Microsoft offer all that? I don’t know. What I do know is that without meaningful technical, commercial, and go-to-market advantages, no serious manufacturer is going to trade Windows flexibility or SteamOS simplicity for a gaming-specific Microsoft OS.
The Store suffers from the same calculus. If a multi-device gaming OS is going to matter, it has to matter on PC. The Xbox Store has to be credible first on catalog, reliability, refunds, reviews, ownership, and developer economics. Then it has to be different: cross-buy, cross-device identity, and a simpler publishing path.
Gamescom delivered a small but useful test. Play Anywhere got The Witcher 3 Remastered via the Xbox Store. A second test released Battle.net with a Diablo IV promotion. One tests the cross-device entitlements promise. The other tests whether a Microsoft-owned PC storefront can lure an outside publisher.
Both make sense. Unless Xbox Store and Battle.net share identity, entitlements, discovery, or economics, Microsoft is running two experiments against Steam rather than building one platform.
Does anyone think that platform is here today?
I wouldn’t try to keep Steam off Xbox hardware. I would try to make Xbox the best living room and handheld environment for Steam, Xbox, Epic, GOG, Battle.net, Game Pass, and direct developer relationships to coexist.
That’s the architecture. In Part II we’ll look at the uncomfortable math of who pays for that openness, and what happens should this platform strategy not pan out.
Microsoft’s new studio map now makes the question unavoidable: who pays for an open Xbox platform, what must Microsoft own, and who has the authority to force the tradeoffs?
Part II — Who pays for openness?
Part I was about making the case for a credible console anchor and for a platform that extends beyond the console. Part II is about who pays for this openness, what pieces Microsoft genuinely must own, and what they do if the platform no longer looks worth the capital.
Revenue-wise, mobile is already working. AppMagic’s estimates of reported 2025 in-app spending are above $1.7 billion, including $1.3 billion from King and more than $100 million from Minecraft.
King has a massive casual audience, earned mobile operating skill, and presence well beyond the traditional Xbox customer base. It’s a profitable mobile business. No strategic poetry needed here.
It’s more difficult to call mobile an Xbox platform. The store that was promised for July 2024 didn’t appear, and Sharma could only say in April that “the idea of an Xbox mobile store is not dead.” Apple and Google are major hurdles. So too is product-market fit. Moving a console game to a phone doesn’t inspire a mobile audience.
Mobile helps Microsoft Gaming’s bottom line, but it doesn’t necessarily flow benefits to Xbox and its platform ambitions. Dressing up Candy Crush in Xbox green would likely be bad for both brands. King should keep the name that its customers recognize while bringing cash, presence, and skill to the broader business.
Minecraft is more interesting. It has a long-lasting multiplatform revenue stream, one of the best-loved entertainment brands in the world and is growing into a creator and commerce platform alongside Roblox and Fortnite.
Sharma has promised to invest in Minecraft “more than ever before” while expanding tools that help people “create, share, build audiences, and earn.” Microsoft IDs already play critical roles in the cross-platform experience, and the Minecraft Partner Program enables approved creators to sell work through Marketplace.
The question for Xbox is whether these capabilities make the wider platform more valuable to creators and players outside of Minecraft. They may simply be valuable inside Minecraft’s kingdom. Microsoft owns the business and earns the money either way; the relevance of Xbox doesn’t increase just because kids are building their latest Minecraft creations on an iPad.
Mojang and King now report directly to Sharma because they have “increasingly become platforms.” Good. Will their platform thinking reshape Xbox, or are they more likely to just help improve the Microsoft Gaming margin? The answer will tell us what Microsoft is really trying to do.
Transformation, or retreat?
On September 22, Microsoft made its operating model more explicit.
Activision will oversee Rare and World’s Edge and build the next Halo through a purpose-built team. Bethesda adds Obsidian, including Grounded and its new Fallout collaboration. King adds Microsoft Casual Games. Playground and Turn 10 become one studio focused on Forza and Fable.
Compulsion and Double Fine are independent with their IP, back catalogs, and runway funding. Undead Labs has now completed its transition and will release State of Decay 3 day one on Game Pass with a new publisher. Two attempts to sell Ninja Theory fell through, leaving the studio in consultation over a proposed closure. Arkane remains under consultation through the end of the year.
Read charitably, this is overdue operating discipline: fewer business units, clearer franchise ownership, and proven publishers matched with adjacent capabilities. Read less charitably, Xbox Game Studios is being hollowed out while the acquired publishing empires become the operating company. Either way, the test is the same. Lower costs and better franchise throughput may improve Microsoft Gaming. They don’t automatically strengthen Xbox.
Microsoft doesn’t need dozens of owned studios to close the gap. When we put together the original Xbox lineup, we either bought teams or partnered with or funded them. The goal was to identify the best games, to make sure they could be built, and to acquire enough of them to make a platform.
This playbook still works: selective publishing deals, co-development, funded exclusives, well-designed incentives. It just isn’t free. Moving a project out of Microsoft changes who carries the studio weight; it doesn’t eliminate the cost of the content or the hit risk. If savings are banked instead of put back into distinctive Xbox games, the brand will continue to hollow out.
Thomas Tippl, the former Activision COO, offers the strongest challenge to my first essay. He thinks Xbox is too far down the priority chain inside Microsoft and should be “on its own feet.” I wouldn’t rule this out. A spinout would force focus and accountability, but it would also mean giving up some of Microsoft’s ability to subsidize the transition and would force Xbox to pay explicitly for Windows and Azure.
But “spin out Xbox” isn’t yet a transaction plan. Microsoft is going to have to draw a line around what franchises, studios, store entitlements, subscription obligations, shared services, cash flows, and liabilities go where. Until someone can articulate what leaves Microsoft and enters this new company, it remains a thought experiment.
Microsoft should use this reboot opportunity to establish a truly independent Xbox P&L inside the parent before even considering a spinout. The test is whether Chiang gets a P&L with teeth, not just a P&L for which she is held accountable when the numbers don’t come through. She needs a multiyear capital commitment.
There are reasons this group might work. Chiang now owns the end-to-end Xbox P&L and understands Xbox Live and Mojang. Asha comes in with fewer loyalties to the status quo. Ball brings an outsider’s perspective. The old prophecy had plenty of believers. What it didn’t have was an accountable P&L with clear decision rights across the stack.
The new studio map gives that P&L its first real test. Activision can rationally want Halo everywhere while Xbox can rationally need signature games that make its platform matter. King can optimize a huge mobile business without contributing much to Xbox identity or ownership. Bethesda can maximize Fallout while the Xbox Store remains an afterthought. A single P&L matters only if it can decide which objective wins, and compensate the group asked to serve the larger strategy.
So watch where the savings go. Reinvestment in Helix, the PC store, a mobile platform services infrastructure, developer tools, and a selective lineup of signature games would signal that Microsoft is still trying to build a platform. Cash extraction combined with ongoing asset sales would look like managed retreat regardless of what the memo says.
If things end up breaking this way, my bet is on more piecemeal sales rather than a full-on sale of Xbox. A company that is principally interested in profitable entertainment franchises doesn’t particularly need an Xbox operating system (and, eventually, perhaps doesn’t need Xbox hardware). Would that still be a Microsoft gaming business worth calling Xbox?
I don’t think so.
The choice I would make
If I got to write the doctrine, I’d pick platform. Stabilize the console, make it the reference device, but don’t confuse the two. Turn Xbox Mode into a coherent XboxOS on Windows machines and third-party devices. Make ownership and developer reach the reason people pick your brand. The Store, first-party games, Game Pass, the cloud, mobile – all to support that promise. And yes, give your fans a bit of hope.
An Xbox that’s only good for launching Steam is like subsidizing Valve’s cash register.
Microsoft has lots of places it can collect money: Store transactions, first-party sales, Game Pass, PlayFab and Azure usage, identity and entitlement services, advertising, OEM licensing. Done. Now the question is which layer you’re subsidizing, what you’re charging for, and what you’re going to own because players and devs genuinely like it better. You can’t subsidize a tollbooth on an open system.
My approach would be straightforward: subsidize what makes the network larger and better; charge for what Xbox brings to the table that’s differentiated; own what few services are needed to make the whole experience trustworthy and portable.
The reason to pick Xbox: exclusive, signature games; cross-device ownership; developer reach. Buy it once and you own the game on console and PC. Sign in and your library, game progress, achievements, and friends list await. Give devs one path to Xbox console players and the Windows installed base. It’s a brand promise people can understand.
What Microsoft has to own and be good at: identity, entitlements, commerce, community. These are what make cross-device ownership and engagement possible. They’re the scarce layer everything else sits on.
What Microsoft gets paid for: Store transactions, first-party sales, Game Pass, OEM licensing. These businesses can pay for the platform. (Enumerating them doesn’t answer the question. That’s the strategic question Microsoft has to answer.) Game Pass should be an enabler of access and discovery; it can’t be the sun around which everything else orbits.
The plumbing: PlayFab, Azure, DirectX, dev tools, cloud saves, streaming, advertising. Some of these will earn nice usage revenue at scale. Wherever possible, Xbox should subsidize dev access. The plumbing should eliminate friction, not create another tollbooth. The reason to pick Xbox should be higher in the stack.
And there it is: what to own, what to subsidize, what to charge for, and in which order. Easy to type. Mind-bogglingly hard to do.
Think about the fights Chiang must fight: Xbox protects Store transactions or does it welcome Steam? Is Helix subsidized like a console or priced like a PC? Does Xbox Mode serve Windows or does Windows have to serve Xbox Mode? Which P&L gets credit when a player uses Microsoft identity, PlayFab, Azure?
Microsoft has built some of the most successful platforms in the world. What it has also mastered is making one division’s opportunity look like another division’s problem.
Every Windows advantage that comes to Xbox carries an organizational veto point. Chiang has to be able to have these conversations with Windows, the Store, Azure. Asha must hold enough capital to get through this transition patchwork of ugly margins. Without both, Microsoft will play this same game with a different cast.
What success looks like
We don’t have to wait 12 to 24 months for the first evidence. Does Microsoft explain what Halo is supposed to do for the platform, not merely who develops it? Do the savings flow into Helix, Xbox Mode, the Store and entitlements, developer tools, and signature content? Do the new reporting lines actually shorten decisions?
Some ‘try a few things and see what sticks’ experimentation is inevitable when you’re jumping from a burning platform. Experiments become a strategy when they have owners, deadlines, metrics, and a reason to stop. We should see evidence well before FY 2030. Within 12 to 24 months, Microsoft should be able to point to third-party XboxOS devices; rising Play Anywhere and PC Store usage; Disc to Digital migrating into Helix; genuine progress in Minecraft’s creator economy; faster publishing for developers; a decision about the relationship between Battle.net and the Xbox Store; and an exclusivity policy that survives more than one reporting period.
The ways we’ll know it’s failed: Helix is premium-priced and there’s no clear reason to pick it; Xbox Mode is mainly a launcher for stores from other companies; another mobile delay; getting casket-ready on crown-jewel assets; savings funneling up to Microsoft rather than back into the platform.
A managed retreat can be a perfectly valid corporate decision, even if it’d make me sad. Microsoft could decide to harvest the profitable franchises, downplay its hardware exposure, and let the Xbox platform contract. In that case, they should just come out and say it. The worst outcome is to promise a platform and run a harvest portfolio.
Microsoft is a platform company. If they can’t make the old Xbox prophecy into a platform that players, devs, and OEMs pick because they want something that makes their lives better, divestiture won’t look like a thought experiment anymore. It’ll look like the destination.
Jon Kimmich is CEO of Software Illuminati, a games consultancy. He worked at Microsoft for 16 years, including during the conception and launch of Xbox.