Finn Brooks: Clara, hey — I want to start with something weird, because I've been reading this Asha Sharma memo and there's a line in it that I genuinely cannot get out of my head, and it has nothing to do with the layoff number.
Clara Bennett: Go.
Finn Brooks: She writes — and this is the Xbox CEO, appointed February 2026, in an internal memo — she writes that the business is 'not healthy.' Not 'facing headwinds.' Not 'undergoing transformation.' Not healthy. That is a clinical word. And then she attaches a number to it: lost 64 cents for every dollar invested.
Clara Bennett: Which is her way of making it unforgettable. You put that number in writing and it's the only thing anyone quotes.
Finn Brooks: Right — but the backdrop is 4,800 total jobs cut at Microsoft, with Xbox absorbing 3,200 of them. That's not a supporting character in this story, that's almost 20% of Xbox's entire global workforce.
Clara Bennett: And the cuts extend through fiscal year 2027, so this isn't a single announcement — it's a restructuring timeline. Meanwhile Amy Coleman, Microsoft's Chief People Officer, is telling the broader workforce this is about responding to AI-driven industry change.
Finn Brooks: Which feels like... I want to believe it, but walk me through it, because Xbox's install base was lagging and Game Pass growth was slower than expected way before AI became the thing everyone points to.
Clara Bennett: That's exactly what we need to untangle — whether the AI framing is the actual driver here, or whether it's the more palatable story. Because those are different problems with very different fixes.
Finn Brooks: Okay but here's where it breaks for me — the AI framing, like, it's not wrong exactly, but it's doing a lot of lifting for something that was already on the floor. The Activision Blizzard deal closed in 2024, $68.7 billion, and that's supposed to be the thing that saves Xbox. And now three years later they're cutting 20% of the division.
Clara Bennett: Think of it this way. Imagine a restaurant that's been losing money on every meal for five years. And then the owner announces they're closing the kitchen because delivery apps changed the industry. The delivery apps are real. But the kitchen was already underwater.
Finn Brooks: That's — yeah. That's exactly it.
Clara Bennett: Xbox's margins were running 3 to 10 times lower than comparable platform and publishing competitors. That's not an AI problem. Game Pass growth came in slower than projected — that's a pre-AI story. The install base lagged. The cost structure was bloated. Those are five-year-old failures dressed in new language.
Finn Brooks: Wait — 3 to 10 times lower? That's not slightly behind, that is a completely different business.
Clara Bennett: Which is why I'd push back on calling this an AI pivot at all. Microsoft is spending $2.5 billion to embed 6,000 engineers inside enterprise clients for AI adoption — that's real, that's happening — but that capital didn't cause Xbox's unit economics to collapse. It just made the contrast impossible to ignore.
Finn Brooks: So the news isn't that AI killed Xbox. The news is that Sharma actually wrote the number down. Like, 64 cents lost per dollar invested — with her name on the memo.
Clara Bennett: That's the signal, yes. The specificity is what's new. A number that precise in writing is not crisis communication — it's a controlled admission that forecloses the softer story.
Finn Brooks: So she's basically saying — we're not blaming the market, we're not blaming the AI wave, we know exactly how broken this was and now you do too.
Clara Bennett: But here's the part that complicates that—the circulating take right now is that Sharma's doing something brave, that radical honesty is what real leadership looks like. And I want to test that, because I'm not sure it holds.
Finn Brooks: No, I don't buy the brave read. Not fully.
Clara Bennett: Walk me through why.
Finn Brooks: Okay — Ninja Theory, Undead Labs, Compulsion Games, Double Fine Productions. All four being transitioned to outside management. That is happening right now, to real studios, real teams. And the CEO's memo is public. So if you are a developer at any of those four places, that memo isn't leadership transparency — it's your CEO announcing on the record that the division you work for is broken. That's the plane's engine light coming on while you're still boarding.
Clara Bennett: That's fair — and I feel the force of it. But the counter is: without naming the problem exactly, any reset loses credibility. Vague optimism from a CEO in this situation would've been worse. Developers can smell a managed narrative.
Finn Brooks: Maybe — actually, wait, here's where I keep getting stuck. It's the timing. Arkane Studios France isn't being transitioned, they're being sold off entirely. And then in the same restructuring, Helen Chiang gets promoted to a brand new COO role, and Mojang and King get pulled directly under Sharma. That's not a confidence-building org chart — that's a company retreating to the franchises it knows won't die. Minecraft isn't going anywhere. Candy Crush isn't going anywhere. Everything else? Less certain.
Clara Bennett: So the memo says 'we're being honest' and the org chart says 'we're hedging.'
Finn Brooks: Those two things don't fully cohere. And that's what gets me about the 'strong leadership' framing — the honesty might be real, but the structure underneath it reads like a contraction, not a rebuild.
Clara Bennett: And the question I'd want answered — which we don't have the answer to yet — is whether any of these divested studios actually survive independently, and whether a flatter org under Sharma ships games faster or just ships fewer of them. That's the part of this story that hasn't been written.
Finn Brooks: Right — and I think that answer tells us whether this is a reset or a slow exit dressed up in restructuring language.
Clara Bennett: And that slow-exit read actually has a historical anchor, because Satya Nadella has done this before — Nokia. Microsoft bought Nokia's devices business, couldn't make the unit economics work, and eventually just walked away. The cloud pivot that followed wasn't a surprise to anyone watching the margin profile. Azure and Office run at 30 to 40 percent gross margins. Gaming doesn't get close.
Finn Brooks: Wait — so you're saying the Activision deal, $68.7 billion, might follow the same playbook as Nokia? That's a massive number to just... quietly back away from.
Clara Bennett: Not necessarily the same ending, but the pattern is there. The acquisition was supposed to deepen the content library enough to justify the cost structure — Call of Duty, Candy Crush, the whole pipeline. Now Ninja Theory, Double Fine, Compulsion, Undead Labs — those are the studios that were meant to fill the gaps Activision Blizzard didn't cover. And they're being spun out.
Finn Brooks: So the bet was: buy Activision, build depth, win Game Pass subscribers. And instead, the content pipeline is actually getting hollower post-acquisition than before it.
Clara Bennett: That's the concrete thing to watch — do Ninja Theory and Double Fine find independent funding, or do they quietly go dark within 18 months? Because if they collapse, that's not restructuring. That's the $68.7 billion bet admitting it didn't buy what it thought it was buying.
Finn Brooks: And nobody's talking about what it means for a developer greenlit under the old structure — I mean, imagine you got the go-ahead on a project in, say, early 2025, you've got a team built, and now your studio is being handed to outside management mid-production. That project doesn't automatically survive the handoff.
Clara Bennett: Right — and that's where the flatter org question actually bites. Sharma's restructuring is supposed to ship titles faster. But faster only matters if the studios doing the shipping still exist and still have funded projects in the queue.
Finn Brooks: So the two things to actually track: studio survival rate, and whether the first game out of the restructured org ships on a shorter timeline than the pre-cut bloat produced.
Clara Bennett: Those two data points, more than any memo, will tell you if this is a reset or an exit in slow motion.
Finn Brooks: And maybe that's the actual question I'm left with — like, Sharma declared the division not healthy and then removed Ninja Theory, Double Fine, Compulsion, Undead Labs from the internal structure. That's not ambiguous. But Microsoft has never once said gaming is dispensable. They just keep acting in ways that look like... they might believe it is.
Clara Bennett: And they've also never shown that gaming fits the margin model. Azure, Office — those numbers are real. Gaming's numbers are also real. Nobody at Microsoft has explained how those two things coexist long-term.
Finn Brooks: Which means we're just... watching, I guess.
Clara Bennett: Watching to see which one is actually true. That's an uncomfortable place to end, but it's the honest one.
