Almanack.
← Stories
Rule

When the market shifts, willingly cannibalize the locked-in business that made you by shipping your crown jewels everywhere, trading device dependence for a subscription bundle worth far more.

Microsoft

Microsoft is the enterprise software company built on the Windows and Office franchises, re-architected after 2014 by CEO Satya Nadella around cloud and cross-platform subscriptions. Its stories share one through-line: a dominant incumbent choosing to cannibalize the businesses that made it, trading device lock-in for a subscription-and-cloud bundle that turned out to be worth roughly ten times more.

Pivots: inverting a decades-old Windows-first strategy

The problem. For years Microsoft tied everything to the Windows monopoly, and the strategy had stalled the company: when Nadella became CEO in February 2014, Microsoft was valued at just over $300 billion, having lost roughly 30% of its value across the 14 years Steve Ballmer led it, while the market moved to cloud and mobile devices Microsoft did not own. Defending Windows meant withholding Microsoft's crown jewels, Office, from the iPhones and iPads knowledge workers were actually using.

The approach. Nadella made his first public act as CEO the launch of Office for iPad, on March 27, 2014, shipping Office onto Apple's hardware and reframing the company as "mobile-first, cloud-first" rather than device-first. The deeper move was to stop protecting Windows as the center of gravity and go everywhere: iOS, Android, Mac, and Linux running on Azure.

How it solved it. By 2024 Microsoft's market capitalization had reached roughly $3 trillion, about a tenfold increase under Nadella, making it briefly the most valuable public company in the US. Freeing Office and Azure from the Windows tether unlocked the growth that a Windows-only posture had capped.

Recurring Revenue: from one-time licenses to recurring subscriptions

The problem. Microsoft's traditional model sold perpetual software licenses, one-time Office and Windows Server purchases that produced lumpy, upgrade-cycle-dependent revenue and gave customers no reason to keep paying. That model also punished the company for widening its reach, since a copy of Office on a Mac or iPad was a lost Windows sale under the old logic.

The approach. Microsoft shifted its crown-jewel products to recurring subscriptions, Office 365 and later Microsoft 365, priced per seat per month rather than per license. Because SaaS economics reward maximizing the total addressable market, the subscription model actively encouraged the cross-platform strategy: every additional device was another billable seat, not a threat.

How it solved it. The recurring model scaled and compounded: in the quarter ending June 2024, Office 365 Commercial revenue grew 16% on 7% seat growth, and Microsoft 365 Consumer subscribers rose 10% to 82.5 million. Predictable per-seat revenue replaced the boom-and-bust of license cycles.

Business Model: re-platforming from on-prem servers to Azure cloud

The problem. Microsoft's server business sold software that customers installed and ran in their own data centers, a model being disrupted by cloud infrastructure that Amazon's AWS pioneered. Continuing to sell on-prem Windows Server and SQL Server would have ceded the fastest-growing layer of enterprise computing to a competitor.

The approach. The core change was to the business model itself, not just the product: Microsoft rebuilt its offering as Azure, a consumption-based cloud where customers rent compute and storage, and it embraced running rival technologies (including Linux) on that cloud to win workloads. This deliberately cannibalized lucrative on-prem license sales before challengers could take them.

How it solved it. Azure grew into a top-two cloud alongside AWS; in the quarter ending June 2024, Azure and other cloud services revenue grew 29%, and total Microsoft Cloud revenue hit $36.8 billion, up 21%, at an annualized run rate of $147.2 billion. The self-disruption that killed Kodak-style incumbents became Microsoft's largest growth engine.

Moats: winning by doing everything, so the bundle beats best-of-breed

The problem. In any single category, a focused rival could out-build Microsoft: Slack was a better chat product, Zoom a better video product, specialists a better fit for technical users. A moat built on being the best at one thing was not defensible against category-specific challengers.

The approach. Microsoft leaned on integration rather than point-product excellence, bundling Microsoft 365, Teams, Azure, and identity (Active Directory / Entra) into a suite where the whole is worth more than the parts, especially to the non-technical knowledge workers who are most of the market. As Ben Thompson observed, Microsoft wins not by being best at any one thing but by doing everything, even mediocrely, well enough inside one bundle.

How it solved it. Bundling Teams into existing Office 365 seats at no extra charge blunted Slack despite Slack's product lead; Teams reached 75 million daily active users by 2020, adding 31 million in a single month during the pandemic. The advantage was strong enough that regulators intervened: Slack filed an EU antitrust complaint in 2020, and Microsoft was ultimately forced to unbundle Teams from Office globally in 2024.

Network Effect: file formats and org-wide collaboration that get stickier with scale

The problem. A productivity suite is only as valuable as the people you can exchange work with; a lone user of a rival format is stranded when everyone else sends .docx and .xlsx files. Microsoft needed the value of its suite to rise with each additional organization on it, not just with its own feature list.

The approach. Office's file formats and collaboration created a classic standards network effect: because nearly everyone uses Office documents, each new user has to as well, and each org-wide rollout raises the value of staying in the suite. Microsoft extended that same dynamic into real-time communication with Teams, whose value compounds as an entire company standardizes on it for chat, meetings, and calls.

How it solved it. The network effect made the bundle self-reinforcing: once an organization ran Teams company-wide, the switching cost was not one product but the whole interconnected fabric of documents, identity, and communication. Teams' surge to 75 million daily active users in 2020 reflected org-wide adoption, where one company's rollout pulled its employees, and their outside collaborators, deeper into the Microsoft standard.