Microsoft is the chief target of accusations of unfair competition, buggy software, and general conspiracy theories. The company could learn a few tricks from an old dog like IBM.

Back in 1987, I attended a large meeting at IBM's newly built Norwegian headquarters, where the company presented what would be their main software platform — OS/2, particularly OS/2 Extended Edition. This was a 32-bit operating system with a database system included, a long-awaited industrial-strength platform for serious, distributed enterprise application development. The presentation was impressive, the company no less so (with more than 80 percent market share in the then-dominant mainframe market), and trade journal articles following the event bemoaned the death of software innovation, now that IBM was taking over everything with what would surely take over.

That didn't happen, of course. There are many reasons for this: Bill Gates flawlessly executed his "take over the world with Windows" strategy, laid down in 1984 (as reported in the January issue of Fortune, when Microsoft had revenues of $50m). OS/2 was more machine-heavy and less functional (not to mention too tightly integrated with IBM's PS/2 line of PCs) than it needed to be. But mostly, IBM shot itself in one of its many size-11 black-shoe feet, being handicapped by its own size and long dominance. For one thing, the PC market was a minor market to IBM. Secondly, IBM's employees were more focused on IBM's internal measures than what the customers wanted. In the PBS TV series based on Bob Cringely's Accidental Empires, there is a sequence where Steve Ballmer describes the experience of co-developing OS/2 with IBM, how the whole thing became a fantastic clash of corporate culture, with Microsoft having the small-company attitude of getting things done, and IBM being focused on internal measures — primarily KLoCs, that is thousands of lines of code as a measure of programmer productivity. "All they cared about," raves Ballmer, "was KLoCs and KLoCs and KLoCs." IBM, apparently, did not care whether the code was any good as long as there was a lot of it.

Anyone with experience of IBM in those days would echo Ballmer's observations. IBM's ability to export their bureaucracy to their customers was outstanding: I remember needing some manuals for our mainframe computer back in those days, and ended up spending two hours with a corporate drone trying to force me through the IBM computer manual course. Yes, there really was a course in how to order documentation from IBM, with certain kinds of manuals being ordered from Oslo, some from Paris, and some from the States. After two hours I told the person that if I we needed any manuals, we would call her, and she could figure out where to get them.

Consequently, the good salespeople in IBM (and there were many) saw it as their job to fight their own bureaucracy on behalf of their customers, a position that was officially sanctioned by top management but tacitly sabotaged by everyone else. An elaborate system of measures was set up to counter this, with rewards taking on a life of their own, spawning a culture where IBMers were constantly talking about who would be on the executive committee and when they would make the 100 club, even discussing these internal machinations with their clients as if we cared.

Now, that was IBM in the old days. In the early nineties, the firm suffered grievous setbacks, had to abandon their life-time employment policy, lost their dominance in the computing market in a classical disruptive technology plot — and, having enough money to survive for a while and the good sense to hire a manager from outside, reinvented themselves as a server-and-service company with, of all things, a focus on open source software. That IBM employees one day should be arrested for painting "Peace, Love and Linux" on San Francisco sidewalks was not something you would believe if you were one of their customers in the 1980s, to put it mildly.

So, how about Microsoft?

Microsoft has now enjoyed a market position in personal computers very much like IBM's in mainframes since about 1993. Their core market — operating systems for personal computers and the standard productivity software we know as Microsoft Office — is maturing and commoditizing — and the extent to which its dominance in this market can be transferred to other markets, such as servers, mobile handsets, and intra-organizational collaborative software, is debatable. The company has taken over IBM's position as the chief target of accusations of unfair competition, buggy software, and general conspiracy theories.

Now, Microsoft is not IBM — the culture is not as corporate, top management has coding experience and technical savvy, and the company is smaller and more focused than IBM was in the 1980s. However, there are danger signs.

I recently attended a presentation of a Microsoft strategy for a particular market — worth about $1b — where the company's particular strategy was outlined. And, touch wood, it was just like being back at IBM, with long lists of blue squares with incomprehensible four-letter acronyms, assurances that Microsoft had products or at least standards for every conceivable use, and an implied promise that "nobody has been fired for buying Microsoft." I caught myself looking round for 3270 terminals.

It seems to me that Microsoft's main problem is running out of new markets to enter. There are many markets available that are too small for it, at least too small for a pure product strategy. These markets require knowledge of how to solve the customers' problems, and choosing which products to use after the problem is understood. Microsoft is product-focused rather than solution-focused, and as such will always be enslaved to their main markets, their main users, and their history.

IBM managed to break out of this hole, chiefly through bringing in outside management, changing the organization (and the people in it) and propagating an understanding of the level of the crisis. I think Microsoft is in the same situation, but there are some differences: Microsoft is not in a financial crisis, quite the contrary. It is facing disruptive technology in the shape of Open Source software, primarily Linux, and taking predictable steps to counter it: Sustainable technology evolution in its own product, such as automatic updates and more focus on reliability; image improvement through spending on research and computers in schools and a more accessible organization (through initiatives such as Channel 9); and continued, if less visible, attempts at muddying the waters for Linux through discrete support for SCO.

Microsoft's problem, however, is its own success: Computers are getting good enough for most people, the new markets are not driven by network externalities (and if they are, the operating system is not the layer where it manifests itself), and the company is not in an underdog position, because the historic alpha dog, IBM, has gone soft and cuddly on them. Continued functionality extension is great for the super-users who are left, and increased reliability is great for all of us (not that we will pay for it, though.) The last time something similar happened, with the Internet, Bill Gates turned on a dime and embraced the technology that could have killed the company, by aggressively using Microsoft's dominance in two markets — operating systems and applications — to gain a dominant market share in browsers. But their heart is not in it. Internet Explorer has not been functionally updated for years, the focus is on extending MS Windows onto new platforms (servers, handhelds, smartphones, media boxes). Microsoft remains, as Neal Stephenson says in his brilliant essay "In the beginning was the command line," an operating system company.

I think Microsoft needs to learn from IBM, and split itself into a collection of companies that cooperate with well-defined interfaces — technical, organizational and financial — between them. Look at what happened to Lexmark after IBM spun them off. Look at what happened to IBM when they finally got rid of the internecine strife that was the AS/400 people vs. mainframes. Microsoft can hold onto the integration between application and desktop, but I don't think they can hold onto the link between different markets — server and desktop, handheld and desktop, phone and desktop, just to mention a few. They can't because they are Microsoft, and the markets know it. American Airlines spun off SABRE when it became clear that the revenues of their IT division was hampered by the connection to one airline, and prospered for it.

Middle age enters when it becomes clear to you that you are not the person that you want to be, when you realize that the skills that took you to where you are now will not take you further, when you need to switch from increasing your space to tending to what you have. I think Microsoft is entering middle age, whether it wants to or not (and who wants to, or even admit to it happening). Unlike people, however, companies can have youthful parts — and they need to be free to grow.

We'll see what happens.

About the Author Espen Andersen is associate professor with the Department of Strategy at the Norwegian School of Management BI (www.bi.no), and European research director with The Concours Group (www.concoursgroup.com), an international IT and management research and consulting organization. Based in Oslo, Norway, he has done research, spoken, published and consulted on a variety of IT, technology and strategic management issues. In his spare time, he spends time with his family, reads, drinks wine and avoids physical activity. He welcomes visits to his blog at www.espen.com/weblog.