Three years ago, a senior executive involved in Microsoft's Internet operations was offered a job by America Online. He said no, and smiled to himself. "We all were so clueless," he says now, "I remember thinking to myself, [at Microsoft] I'm in the catbird seat, it's only a question of how to use it."
The former Microsoftian can tell this story without embarrassment because he was by no means alone in betting on the wrong horse. America Online then was in the middle of what looked like public relations armageddon as it earned the nickname America On Hold. Companies like Yahoo and Amazon were hardly blips on the screen to anyone but a few analysts. Though Netscape was still considered the Internet company, virtually everyone involved with the Internet at the time believed that if any company would dominate in the businesses that the Net was creating, it would be Microsoft.
Last week, Microsoft announced its latest Net strategy. The announcement of a new Microsoft strategy is a perennial ritual that lets Microsoft flex its muscles and explain how it will remedy its mistakes. This time, Microsoft announced a whole bevy of initiatives that together might be considered quite a big deal: the spinoff of its online travel agency, Expedia, into a separate public company; a new small business site called bCentral; plans to develop cheap new "Internet appliances" -- computers dedicated to Web access and e-mail and running the Windows CE operating system; a revamp of the keystone MSN.com site. Yet what it all added up to for most observers was a collective yawn of "we've been here before."
The biggest news out of Microsoft's press and analyst conference, in fact, was an off-hand remark by Microsoft president Steve Ballmer that technology stocks -- including Microsoft's own -- were overvalued. That got everyone's attention and sent Microsoft's stock down by almost $5 a share in a few hours.
Watching Microsoft in the 1990s is what watching the Kremlin used to be during the Cold War: an exercise in guessing at the meaning of opaque official communiquis and watching carefully for any changes in the seating plan of the Politburo. (Hmm, Microsoft wise man Nathan Myrhvold seems to be spending a lot of time on his fossil collection ...)
Rarely, however, has anyone tried to answer why Microsoft's Net strategy is so regularly disappointing. The answer has a lot less to do with strategic missteps than with some big facts about Microsoft. First, it's not a Net company in the way that a lot of commentators seem to think it is. Second, unlike many of its new competitors, Microsoft can afford to make mistakes.
The easiest way to sum up the first of these points is that while Microsoft watchers want Microsoft to think of itself as a "Net company," it's not, and doesn't want to be one.
In December 1995, Bill Gates famously announced -- at a conference much like last week's -- that Microsoft, the software industry's supertanker, would put the Net at the center of everything it did. Microsoft chronicler Paul Andrews, in his 1999 book "How the Web Was Won" (the title refers to how Microsoft won the Web, and it is hard to imagine that even the author fully believes it), makes much of the fact that the announcement came on the anniversary of the Pearl Harbor attack, likening it to a declaration of war. "You will hear from us that we're not forming an Internet division," Andrews quotes Gates as saying. "To us, that's like having an electricity division or a software division. The Internet is pervasive in everything we do."
It sounds good, but it doesn't really tell the full story. It is true that all of Microsoft's divisions are now focused on Internet services -- that's true of every software company. But that doesn't mean that all of Microsoft is focused on competing with the new breed of "pure" Net companies like Yahoo and Amazon. In fact, in a 1998 reorganization, Microsoft created something that sounds very much like an Internet division, just not called that by name. Instead, it's called the Consumer and Commerce Group, and it includes all of Microsoft's consumer Internet services, from MSN, the online service provider, to Carpoint, an auto shopping site.
"Microsoft," says Greg Blatnick an analyst at the market research firm Zona Research, "is a bit of a platypus. Parts of a duck and parts of a mammal and parts of something else. Amazingly, they all kind of work together."
The platypus notion is a useful idea to keep in mind when thinking about how Microsoft works. Controlling the Net, and Net commerce with it, is only one of the objectives of a big and fairly unwieldy technology conglomerate.
Here are some numbers: In the period from June 30, 1998, to June 30, 1999, the company's total revenue was about $19.7 billion, according to Microsoft's financial statements. Of that, operating systems accounted for $8.5 billion. Productivity applications -- primarily the expensive and ubiquitous Microsoft Office -- got Microsoft another $8.8 billion. In other words, while Microsoft is clearly interested in developing new lines of business like Net commerce and Net appliances, right now it's the operating system and its shrink-wrapped software that is buttering Microsoft's bread. It's an obvious point, but it's worthwhile to take careful note of it.
When Bill Gates talks about extending Microsoft's franchise to the Net, he doesn't mean what a lot of analysts seem to expect him to mean. They think of new Net businesses -- he thinks at least as much of making Microsoft's software Netcentric. In this, his strategy has been a tremendous success. MSN might not have taken over the Net, but the software business is growing at least as fast as ever.
This leads in to the second reason why Microsoft has failed to dominate the Net: Microsoft, with its huge size, can afford to have many plans, while the most successful players on the Net have one.
When Microsoft embarked on its Internet jihad -- the word is Ballmer's -- Compuserve looked like a Goliath (remember, as the former Microsoft exec said, we were all clueless), America Online looked like an ascendant Net player and the Microsoft Network, a proprietary online service, was Microsoft's biggest bet on the Net. To its credit, Microsoft changed direction fast -- and it did so repeatedly. It moved from creating beautifully produced Net entertainment shows like the defunct Mungo Park (still one of the Net's nicest ghost sites) to creating Net businesses like Expedia and Carpoint. It shifted again to follow the conventional wisdom and launch a mega-portal site, MSN.com. Since the end of 1998, when Gates sent around a well-known memo extolling the virtues of connectivity anytime and anywhere, Microsoft seems to have been moving in still another direction, designing cheap Net access appliances for the living room and trying to get a toehold for Windows CE in cable companies' set-top boxes.
Now compare that to a company like Amazon, the giant Net retailer with which Microsoft shares the state of Washington. You want changes in strategy? Well, in 1997, Amazon, then still just a bookseller, started selling music and videotapes. You could maybe, if you insist, count that as a strategy change. Then Amazon started toys and games. Then electronics. On Thursday, Amazon added still another wing to its emporium, opening up an online mini-mall of specialty shops.
OK, so none of those are big shifts in strategy, and that's the point. When Amazon's Jeff Bezos changes his mind, it's usually about something so deep in the bowels of the company that hardly anyone even notices. (Did you know that Amazon has now moved to keeping more inventory on hand? Well, it did, but it sure isn't news of the caliber of the regular Microsoft "reorgs.") Why is that? Simple: Microsoft has a Net strategy; Amazon (or Yahoo for that matter) is a Net strategy.
Microsoft is not organized around strategies, it's organized around technological know-how. This means, in essence, that Microsoft can barrel into new businesses without a definite idea of how they will work. Often Microsoft knows the technology, but, -- precisely because it is a big company that might be able to build a consumer Net business alone -- it often doesn't establish the partnerships that make it work. (In business-to-business enterprises, like legal services, Microsoft is readier to make key alliances.)
For instance, Microsoft has for two years been trying to enter the bill-payment business through a venture called TransPoint. Bill-payment technology is something that is buried deep in the back offices of financial institutions, but it's important. Bill payment is at the core of a lot of lucrative consumer banking business that Microsoft and a lot of other companies want to get their hands on. But it took a year and a half for Microsoft even to partner with one bank (Citicorp) and most banks are still shying away from cooperating with Microsoft's plans, fearful that Gates, who has singled out banks for criticism as slow and lumbering businesses, wants to take away their paychecks.
Sums up one Microsoft competitor, Ron Martinez, CEO of Brodia.com, a company that has developed an Internet wallet (shorthand for a one-click Internet payment system, something that Microsoft has now set squarely in its sites): "What the Net enables from a technological standpoint doesn't necessarily map to developing a business. Being able to create a Web site that can provide financial services isn't the same as being a financial services company."
There is one Microsoft, and there are lots of start-ups. The start-ups that have ill-conceived strategies go out of business. Microsoft can afford to make lots of mistakes, while new Net businesses can't. Generally, they only get one shot. This survival of the fittest leaves only those businesses that are built around a Net strategy that works.
None of this, of course, means that Microsoft will fail to dominate Internet commerce in the way that many analysts have predicted it would. Think of it as a war of attrition: Microsoft can lose a lot of times, but eventually it will hit on the right path. When that happens, its competitors will not only have to remain nimbler than Microsoft, but they'll have to be big enough to take on the colossus without being outspent, outmarketed and outflanked by Microsoft's dominance in the software world.