Microsoft's Real Monopoly Was Never the One the Government Sued. It Was a Loop Nobody Drew on Purpose.

Open a job listing for almost any office role today and somewhere in the requirements line sits a phrase nobody questions: 'proficient in Microsoft Office.' Not 'a spreadsheet,' not 'a word processor' - the brand, as if it were a literacy. That sentence is the residue of a competition that ended thirty years ago, and the strange thing is the company that won it was never punished for the win that mattered. The world remembers Microsoft losing an antitrust trial. Almost nobody remembers that the trial wasn't about the thing that actually locked us in.

The popular story is that Microsoft bundled Office with Windows, the government called it a monopoly, and the courts slapped it down. Nearly every clause of that is wrong. The 1998 case was about the web browser, not Office. Office was always sold separately. And the loop that genuinely trapped the business world - Windows feeding Office, Office feeding Windows - was never broken up at all. It is still running.

Microsoft owned the floor before it owned the furniture

Here is the detail that rewrites the whole story. By the late 1980s Microsoft already controlled the operating system - MS-DOS ran on more than 80% of PCs.⁸ But it did not control what people actually used those PCs for. The spreadsheet that ran businesses was Lotus 1-2-3. The word processor that wrote their memos was WordPerfect.⁸ Microsoft owned the floor everyone stood on and almost none of the furniture they sat in.

And Office, when it arrived, did not even arrive on Windows. Bill Gates announced the suite at COMDEX on August 1, 1988, but the first version shipped on June 19, 1989 - for Macintosh only. It bundled Word, Excel, PowerPoint and Mail. Office for Windows did not ship until October 1, 1990, more than a year later.¹ The tidy 'Windows-first' legend is backwards. Office was a Mac product that Microsoft brought home only once Windows was ready to receive it.

The loop nobody drew on a whiteboard

This is the mechanism, and it is worth working all the way down, because the popular telling skips the only part that explains the outcome. When the PC world migrated from MS-DOS to Windows, every application had to be rewritten for the new graphical platform. That migration was a reset - a moment when the leading spreadsheet and the leading word processor had to start over on someone else's foundation. And the someone whose foundation it was happened to also make a spreadsheet and a word processor.

During exactly that transition, Office displaced WordPerfect and Lotus 1-2-3.⁷ Novell, which then owned WordPerfect, would later allege that the displacement wasn't only better salesmanship - that Microsoft exploited undocumented Windows API features to give Office a performance advantage rivals literally could not see, let alone match.⁷ Whether or not you accept the strongest version of that charge, the structural point stands: when you own the platform, you know it better than anyone porting onto it. The home team writes the rulebook and plays from it.

Now the loop closes. The more people standardized on Office, the more they needed Windows to run it well. The more they ran Windows, the more natural it was to buy Office rather than fight the platform's favored child. Each product quietly raised the switching cost of the other. This is what antitrust economists, in the same case, named the applications barrier to entry: nobody would write software for a rival OS that had no users, and nobody would adopt a rival OS that had no software.³ Microsoft sat on both sides of that wall.

Then Windows 95 turned the loop into gravity

If you want a single image of the lock-in passing the point of no return, it is the launch of Windows 95 on August 24, 1995. By Microsoft's own count it shipped 7 million copies in its first five weeks and 40 million in its first year.⁶ At that scale the question for a business was no longer 'which suite is best.' It was 'which suite does everyone I exchange files with already use' - and the answer, on a platform that vast, answered itself. The competition stopped being about quality. It became about gravity.

The trial everyone remembers was aimed at the wrong room

So here is the twist that should reframe the entire memory. When the Department of Justice and twenty state attorneys general sued Microsoft on May 18, 1998, they did not go after Office. They went after the web browser. The charge was that Microsoft unlawfully maintained its monopoly by tying Internet Explorer to Windows and stopping PC makers from removing it.⁴ The Sherman Act violations Judge Jackson found were about the browser. The breakup he ordered was about the browser. Office - the product that had actually consummated the lock-in a decade earlier - was never on trial.

Why the browser and not the suite? Because the browser was where the loop was extending next. Microsoft had learned the move in the early 1990s - own the platform, integrate the favored product, let the network do the rest - and it ran the same play against Netscape on the open web. The DOJ wasn't fighting the original lock-in. It was trying to stop the pattern from swallowing the internet too.

And the legal opening was one Microsoft had carved itself. The 1994 consent decree forbade tying 'other Microsoft products' to Windows licenses - but explicitly permitted Microsoft to 'integrate additional features into the operating system.'⁵ So Microsoft's defense wrote itself: Internet Explorer wasn't a product bolted on, it was a feature of Windows. A semantic distinction, drafted into a settlement years before anyone needed it, became the hinge of the most consequential antitrust case of the decade.

Wasn't this just a master plan executed flawlessly?

The honest objection is that this reads too neatly - as if Microsoft drew the flywheel on a whiteboard and ran it. The evidence says the opposite, and that is what makes it instructive. Office wasn't Windows-first; it was a Mac product for over a year.¹ Microsoft owned the OS layer long before it owned applications.⁸ No grand bundling scheme appears in the timeline - the suite was always sold separately, and the courts never touched it. The lock-in wasn't engineered as a single strategy; it emerged, one self-reinforcing turn at a time, from a company that happened to control the platform when the application layer was forced to reset. The second objection is that the courts ultimately let Microsoft off. They did not. The D.C. Circuit affirmed the monopoly findings of fact in 2001; what it vacated was the breakup remedy, sending it back to a new judge.⁴ The monopoly finding stood. The structure survived it intact.

Microsoft spent years and a fortune in legal fees defending a browser - while the loop that actually mattered, the one between the floor and the furniture, hummed along untouched in every office on earth. The government broke nothing that counted. The market couldn't route around a standard everyone had already agreed to without meaning to. That is the quiet genius of an emergent lock-in: no one chose it, so no one can point to the decision to undo. The most powerful moat Microsoft ever dug is the one it never drew, never bundled, and never had to defend in court - and it's still asking job applicants whether they're proficient in it.