Microsoft Didn't Diversify. It Built a Gravity Well One Acquisition at a Time.

Picture a developer in 2026. She writes code in an editor Microsoft makes, stores it in GitHub, which Microsoft owns, ships it onto Azure, which Microsoft runs, and is recruited for her next job through LinkedIn, which Microsoft also owns. She may never touch Windows once all day. And yet Microsoft is paid at every turn — for the place she works, the work she does, the machine she runs it on, and the career she builds from it. None of that is the operating system that made the company. All of it is Microsoft.

The official story is that Microsoft, trapped in a dying PC business, bravely diversified — bought a social network, bought a code repository, built a cloud, and spread its bets. That reading is wrong in the most important way. These were not bets spread sideways for safety. They were laid end to end. Each one was the on-ramp to the next, and the road they all fed onto was the same: the cloud.

The cloud came first, and it was already moving before Nadella

The convenient version says Satya Nadella walked in as CEO in February 2014, declared 'mobile-first, cloud-first,' and snapped a decaying company awake. The convenient version skips a detail he said out loud. On his first day he framed the cloud strategy as carrying over the devices-and-services direction of his predecessor, Steve Ballmer — not breaking from it.⁸ Azure itself had been announced back in 2008 under the codename 'Project Red Dog' and only went commercially live as Windows Azure on February 1, 2010.⁴ The pivot was years old when Nadella inherited it. What he did was sharpen the point and make the cloud the gravity center everything else would be pulled toward.

And it was small. As late as the third quarter of fiscal 2015, Microsoft's entire commercial cloud — Office 365, Azure, and Dynamics CRM Online combined — was running at an annualized $6.3 billion, more than doubling year-over-year but still a rounding error next to Windows and Office.⁵ The interesting question was never whether the cloud could grow. It was how to make every other part of the company a reason to use it.

Why LinkedIn and GitHub were never really about LinkedIn and GitHub

Here is the part the diversification story can't explain. In June 2016 Microsoft agreed to buy LinkedIn at $196 a share — $26.2 billion inclusive of net cash at announcement, recorded in the FY2018 10-K at a completed cost of $27.0 billion.¹² Two years later, in June 2018, it announced GitHub for $7.5 billion in stock, when GitHub had more than 28 million developers.³ On the surface these are unrelated trophies: a professional network and a code host. Underneath, they are the same move made twice. Each one captures a population that the cloud needs and that nobody else owned end to end.

GitHub is where the world's developers already live — and developers are the people who decide which cloud their employer's next application runs on. Microsoft did not buy GitHub to sell GitHub. It bought the habit that precedes a cloud decision, the place the choice gets made before a procurement team ever sees a bill. LinkedIn does the same thing one layer up: it owns professional identity, the graph of who works where and who is hiring whom — the data that feeds the same enterprise relationships Azure and Office 365 are sold into. The pattern is not a hedge. It is a sequence. Own the workflow, own the identity, and the compute follows almost on its own.

The number Microsoft hid for fifteen years

If you want to know what the company truly cared about, watch what it refused to disclose. For most of Azure's life Microsoft buried it inside the 'Intelligent Cloud' segment and never broke out a standalone figure. You could see the growth — by the third quarter of fiscal 2020, Intelligent Cloud was $12.3 billion in a quarter, up 27%, with Azure revenue alone up 59% year-over-year⁶ — but never the raw size of the prize. That secrecy held for fifteen years. It broke only in July 2025, when Nadella revealed for the first time that Azure had passed $75 billion in revenue in fiscal 2025, up 34%.⁷ A company discloses a number when the number has become the story. The cloud the other bets were quietly feeding had become the largest thing in the building.

Isn't this just hindsight dressed up as a plan?

The honest objection is that this is too tidy. A company buys a social network and a code host years apart, the cloud happens to boom, and a writer connects three dots into a line that was never drawn. Maybe LinkedIn was opportunism, GitHub was developer-relations defense, and Azure rode a tide that lifted every cloud vendor. That's fair, and it's partly true — nobody scripted $75 billion in 2016. But notice what the loose-bets theory cannot account for: the deals reinforce each other in exactly one direction, toward the cloud, and never away from it. Microsoft did not buy a media company or a hardware maker or a payments firm to spread its exposure. Every major bet captured a population whose next decision was which cloud to run on, and every one funneled toward the same destination Microsoft refused to disclose. Coincidence produces scatter. This produced a funnel. A strategy you can recognize only in hindsight is still a strategy if every move points the same way.

The proof is in the staying power of what was bought. GitHub had 28 million developers when Microsoft acquired it; the acquisition did not shrink that population, it grew it.³ LinkedIn still owns professional identity. Azure went from a $6.3 billion commercial-cloud sliver to a $75 billion business in roughly a decade.⁵⁷ These are not assets being managed for diversification's sake. They are tributaries, and they all run to the same river.

Microsoft's real expansion was never about leaving Windows behind. It was about discovering that an operating system is just one way to stand under everything someone does — and that you can rebuild that position, in the cloud, by owning the workflow, the identity, and the compute, in that order. The company didn't diversify away from its core. It moved its core to a place no rival could surround, and then bought the roads leading in. The genius wasn't any single deal. It was that each one made the next one cheaper, stickier, and harder to leave.