In 1975, a 20-year old Bill Gates stated a bold ambition that many at the time thought naïve: his fledgling startup would put a Windows computer on every desk and in every home.
Gates, and his co-founder Paul Allen, never quite realized that ambition. They did however, grow "Micro-Soft" from a tiny, underfunded company selling a BASIC interpreter for the Altair 8800 PC to a $22.9 billion company selling operating systems, applications and tools with a 90% share of the microcomputer software market by the year 2000 -- 25 years into Microsoft's history. Close enough.
That year, Microsoft was among the top five most valuable companies with a market cap of $258 billion. Fast forward to fiscal year 2020: The company's market cap is slightly over $1 trillion and it now tops the list of the world's most valuable companies.
One could surmise that Microsoft's trip to the top of the heap was predictable given its position in the fast-growing computer industry over the past two decades. But the journey between the two mountain peaks saw dramatic changes to the company's senior management and bold technology changes that strayed far from those products that made it rich and famous. It also helped that many of its archrivals made strategic missteps opening doors of opportunity.
Microsoft's history marked by leadership changes
There are the more obvious reasons Microsoft remained near or at the top of the most influential companies in high tech: the arrival of Satya Nadella taking over from Steve Ballmer; the subsequent refocusing from proprietary products to open source, as well as making the cloud its first priority.
But as important to the company's success as the right people rising to the top of the company and the right set of priorities, is an often-overlooked factor: Microsoft's slow and steady progress convincing its mammoth user base to buy its core products through long-term, cloud-based subscriptions.
"If you want to point to one thing that's kept the company's revenues growing over the past 20 years it is subscription selling," said Rob Helm, managing vice president of research at Directions On Microsoft, an independent analysis firm in Kirkland, Wash. "It's not very exciting but the company has used all kinds of clever tactics to shift users to this model. And they aren't done yet."
The move to subscription selling, an initiative that originated before Steve Ballmer took over the day-to-day operations of the company, started with its Licensing 6.0 and Software Assurance program in 2002. The program got off to a slow start, mainly because users were unaccustomed to buying their products through long-term licensing contracts, said Al Gillen, group vice president in IDC's software development and open source practice.
"That program wasn't used much at all. Hardly anyone back then used subscriptions to buy software," Gillen said.
But with the arrival of the cloud, Office 365 and Microsoft 365 in particular, longer-term cloud licensing skyrocketed.
"[Microsoft] became very enthusiastic about the cloud because for them, it was yet another way of moving its customers to long-term subscriptions," Helm said.
The biggest obstacle to moving many of its customers to cloud-based subscriptions was Microsoft's own success. For decades of Microsoft's history, the company was wedded to its business model of selling a stack of products that included the operating system, applications and utilities, and signing up major hardware suppliers to sell the stack bundled with their systems. The company grew rich with this model, but by the mid-2000s, trouble was brewing.
Rob HelmManaging vice president of research, Directions On Microsoft
Under Ballmer's reign, the software stack model became outdated with the encroaching age of the cloud and open source software led by a new raft of competitors like AWS and Google. Nadella saw this and knew it was time to accelerate the company's cloud-based subscription business.
So, although Ballmer famously monkey-danced his way across a stage shouting "developers, developers, developers" at a Microsoft 25th anniversary event in 2000, it was Nadella who knew what Windows and non-Windows developers wanted in the age of the cloud -- and how to go about enlisting their cooperation.
"Satya decisively moved away from the full stack model and went to an IaaS model more resembling that of AWS," Helm said. "This pivot allowed them to deliver cloud services much faster than it could have otherwise. But they could not have done this without cutting Windows adrift from the rest of the stack," he said.
Open source fork in the road
Microsoft's pivot to support Linux in 2018 and open source happened "just in time," according to Gillen. While the company had open source projects underway as far back as the 2004 to 2005 timeframe, including a project to move SQL Server to Linux, it took a long while before open source was accepted across Microsoft's development groups, according to Gillen.
"A developer [inside Microsoft] once told me he would show up for meetings and the Windows guys would look at him and say, 'We don't want you at this meeting, you're the open source guy,'' Gillen said. "Frankly, looking back they were lucky to make the transition at all."
What made the transition to long-term cloud subscriptions easier for Microsoft and its users was the combination of cloud and Linux vendors who already had such licensing in play, according to Gillen.
"It became more palatable to users because they were getting more exposure to it from a variety of sources," he said.
Microsoft has so fully embraced Linux, not just by delivering Linux compatible products and cloud-based services, but through acquisitions such as GitHub, the world's largest open source repository, it is becoming hard to remember a time in Microsoft's history when the company was despised by the open source community.
"If you are a 28-year-old programmer today, you aren't aware of a time Microsoft was hated by the Linux world," Gillen said. "But I'd argue now that Microsoft is as invested in open source software as any large cloud vendor."
Microsoft cloud subscriptions start with desktops
One advantage Microsoft had over its more traditional competitors in moving to a SaaS-based subscription model was the fact that it started by moving desktop applications to the cloud instead of server-based applications that companies like IBM, Oracle and SAP were faced with.
"Microsoft's desktop software is an easier conversion to the cloud and to move into a subscription model," said Geoff Woollacott, senior strategy consultant and principal analyst at Technology Business Research Inc. "The degree of difficulty with desktops compared to an on-prem, server-based database that has to be changed to a microservices, subscription-based model is much harder."
While Microsoft has downplayed the strategic importance of Windows in favor of Azure and cloud-based open source offerings, analysts believe the venerable operating system's life will extend well into the future. IDC's Gillen estimated that the Windows desktop and server franchise is likely still in excess of $20 billion a year, which is more than 15% of the company's overall revenues of $125.8 billion for fiscal 2019 ended June 30.
"Large installed bases have longevity that goes far beyond what most want to have them. Just look at mainframes as an example," Gillen said. "I'd argue that 60% to 70% of Windows apps are going to be around 10 and 15 years from now, which means Windows doesn't go away."
And those Windows apps are all being infused with AI capabilities, in the Azure cloud, on servers and desktops. Microsoft is also making it easier for developers to create AI-infused applications, with tools like AI Builder for PowerApps.
Microsoft's AI, quantum computing future in focus
While Gillen and other analysts are hesitant to predict the future of Windows and its applications in 20 years from now, most believe the company will spend a generous amount of time focused on quantum computing.
At its recent Ignite conference, Nadella and other Microsoft executives made it clear they'll have a deep commitment to quantum computing over the next couple of decades. The company delivered its first quantum software a couple of years ago and, surprisingly, is developing a processor capable of running quantum software with plans to deliver other quantum hardware components.
Over the coming years, Microsoft plans to build quantum systems to solve a wide range of issues from complex problems those doing advanced scientific enterprises face, said Julie Love, senior director of Quantum Computing at Microsoft.
"To realize that promise we will need a full system that scales with hundreds and thousands of qubits. It's why we have this long-term deep development effort in our labs across the globe," she said.
Microsoft's first steps toward introducing quantum technology for enterprise use will be to combine quantum algorithms with classical algorithms to improve speed and performance and introduce new capabilities into existing systems. The company is already beta testing this approach with Case Western University, where quantum algorithms have tripled the performance of MRI machines.
"Quantum is going to be a hybrid architecture working alongside classical [architectures]," Love said. "If you look at the larger opportunities down the road, the most promise [for quantum computing] is in chemistry, science, optimization and machine learning."
Unlike Bill Gates, Satya Nadella isn't promising a quantum computer on every desktop and in every home by 2040. But you can assume that as long as Gates has an association with the company, it will make an earnest effort to put Microsoft software on those desktops that do.