Microsoft has achieved something fairly unusual for a company worth trillions of dollars: it is both enormously important and, to most people, rather boring.
Ask someone what Microsoft does and the answer will probably involve Word, Excel, Windows and perhaps a complaint about Teams. Yet the company now sits behind a much wider part of the technology stack. It provides email, cybersecurity, cloud computing, databases, developer tools and business software, often to the same customer. Much of it is the digital plumbing of modern business: rarely noticed when it works, immediately noticed when it does not.
There is an irony here. The more successful Microsoft has become, the less visible much of its business is. Nobody arrives at work particularly excited that their employer uses Azure, Active Directory or SQL Server. They simply expect the email to arrive, the spreadsheet to open and the system not to be hacked. That invisibility is partly a consequence of how deeply Microsoft has embedded itself across the corporate technology stack. Its advantage increasingly comes not from every product being clearly better than the competition, but from the fact that so many sit alongside one another and are already woven into the way businesses operate.
This is not a new strategy. Microsoft began by selling software for personal computers and quickly realised the value of becoming the common platform. Its relationship with IBM in the early 1980s was crucial. Microsoft supplied the operating system for the IBM PC, while retaining the ability to license it to other manufacturers. As IBM-compatible PCs spread, hardware companies competed fiercely while Microsoft’s software became the layer they had in common.
Companies such as Compaq could build cheaper PCs, but they still needed an operating system. Microsoft licensed DOS on a per-machine basis and benefited from the growth of the wider PC market, regardless of which manufacturer sold the computer. A similar model later emerged in enterprise software. Through the 1990s and 2000s, Microsoft built products around Windows and Office including Exchange, SQL Server and Active Directory. These became increasingly interconnected. Employees used Outlook, Office and Windows, while IT departments relied on Microsoft for identity, security and data. Once several parts of the stack were in place, adding another Microsoft product was relatively painless. Ripping the whole thing out was rather less appealing.
Microsoft also changed how it charged for software. Enterprise agreements bundled products into multi-year contracts, moving the business away from relying on customers buying the latest version of Windows or Office every few years. By 2007, more than half of Microsoft’s revenue was estimated to come through annual or multi-year enterprise agreements.
Microsoft also had something newer competitors had to build from scratch.
That installed base became particularly valuable when computing moved to the cloud. Azure (Microsoft’s cloud computing platform) did not require customers to abandon their existing systems overnight. Companies already using Microsoft software could move gradually, connecting on-premise infrastructure with the cloud. Microsoft also had something newer competitors had to build from scratch: decades-long relationships with corporate IT departments.
This remains one of Azure’s important advantages. Microsoft does not need to win each technology cycle from a standing start. It can introduce new products to customers already using several others.
AI is the latest example. Microsoft is spending heavily on data centres and computing capacity, but the more interesting question is what it can do with that capacity once it exists. Copilot can sit inside Word, Excel, Teams, GitHub, security products and Dynamics. For a customer, adding AI to software already used across the organisation is easier than buying a separate product and integrating it themselves.
GitHub Copilot gives some indication of how this can work. By the third quarter of Microsoft’s 2026 financial year, nearly 140,000 organisations were using the product, while enterprise subscribers had almost tripled year-on-year. Azure AI Foundry (a platform for designing and building AI tools and agents) is trying to do something similar for companies building AI applications, while keeping that activity connected to Microsoft’s cloud, data and security products. The attraction is straightforward: Microsoft can sell more to customers it already has. This is what it has done with Microsoft 365, where customers have moved into more expensive packages containing additional security, compliance and communications tools. AI potentially adds another layer.
There is, however, an important difference. Traditional software was extraordinarily capital light. Once Office had been developed, supplying another customer cost very little. AI requires graphics processing units, data centres, networking equipment and vast amounts of power. If Microsoft builds too much capacity, or demand develops more slowly than expected, returns could disappoint.
Microsoft has already navigated the move from PCs to enterprise software and then to the cloud. AI may be the next transition, but it is a much more expensive one. The question is whether Microsoft can again use the position it has spent decades building to turn the next technological shift into another part of the Microsoft ecosystem.
Please note that the value of securities and the income from them may go down as well as up and you may not receive back all the money you invest. Past performance is not a reliable indicator of future results. Any views expressed are those of the author.
