“Bill Gates has no strategy for Microsoft!”
This was the cry of the business and computing press in the late 1980s. It seemed absolutely accurate.
Except that it wasn’t.
Bill Gates knew something us idiots in the computing world didn’t. Here’s the story.
Back in the 1980s, Microsoft’s core product was the MS-DOS operating system, a massive cash cow that seemed to just print money. Any company in the world would have loved to have had their market domination and it allowed Microsoft to create other products like Word and Excel, and wait out the problems that were legion in their initial releases.
Yes, MS-DOS was the Master of personal computing.
But it looked like more advanced solutions were inevitable. The problem was that no one seemed to know which direction would pan out, and not just “markets being what they are”. The underlying technology was up for grabs.
Servers were becoming cheaper. After years of talk about “the office network”, it seemed that might actually work and servers come into small offices.
There was also the higher end microcomputer market, where the work stations and servers did “real” computing tasks. Companies like DEC and Sun had their market niche locked up with their own operating systems and even hardware.
IBM continued to dominate the personal computing space. This made Microsoft’s deal to provide the operating system a prescient move, but other companies, notably Compaq and Michael Dell’s operation, were creating some machines that were finally getting large corporation’s interest. And Compaq even had its own DOS.
There was even a version of Unix that ran on the X86 architecture, created by the company SCO. Many Unix fanatics were sure that it was the wave of the future because Unix allowed computer users to do so much more than DOS.
Given this situation in the market, what should Microsoft do?
What Gates did was hedge his bets.
He had the cash as a result of DOS’s primacy on the PC, and the market support as the genius founder who had yet to slip up. He chose to enter all the directions at once.
To prepare for Macintosh’s possible supremacy, Microsoft developed a suite of applications for Mac OS and became the world’s largest producer of Mac software.
Looking to the burgeoning server market, it worked with IBM on a server OS (OS/2). But it also developed it’s own “workstation” OS (Windows NT) that could be used on servers if the IBM deal didn’t pay off.
For PCs, MS created a horrible multitasking shell that ran on top of DOS (Windows) and even bought up a sizable chunk of SCO, the Unix on x86 company.
Bill Gates, the press accused, had no sense of direction because he wouldn’t pick one and go that way. He was going lots of ways all at once.
But Bill Gates wasn’t equivocating; he was doing his job as CEO.
The future of personal computing and even corporate computing was absolutely unclear at the time. It turned out that hedging his bets on IBM by working on a “big Windows” paid off and the successors to Windows NT now run on most business servers and desktops. But back then it was murky. It was uncertain. It was strategically uncertain. To make the right bet he needed to manage a portfolio of options.
It’s not the job of lower levels to worry about this. Their jobs were to execute their given areas, creating great software in their areas. Yes, it helped to be on the team that ended up winning rather than MS-DOS at that point, just as the people who were on Jobs’s pet Macintosh went on to be famous while the people on the Apple II-e work at the same time probably just got fired. But in the end, their work was not about managing the uncertainty over what operating system would win but to create the best software in their group.
The amount of strategic uncertainty you manage suddenly jumps when you get to the executive level. It’s probably the easiest way to differentiate between the strategic layer of the corporation and the execution layer. According to Elliott Jaques, Wilfred Brown, Ralph Rowbottom, Gillian Stamp, D. John Issac and others, there is a big jump between the two, a discontinuity in the thinking required. Raynor adds to it by describing how much more strategic uncertainty becomes salient.
We need a better working definition of “strategy”. In my strategic consulting with The Manasclerk Company’s executive clients, I have found a couple of simple distinctions to fundamentally change the conversation around strategy and execution. Tomorrow, let’s take a look at them.
Until then, see Michael Raynor talking about the Strategy Paradox at the Darden School in 2008. (YouTube, ~1hr) It’s also one of the best discourses about Elliott Jaques’s work and what is really going on in these levels.