THE MANY ARMS OF MICROSOFT
Software Giant Reaches for New Markets, Partners, Business Models

by Denise Caruso and Janice Maloney

c. 1994 Technology & Media

If software is the new lingua franca, then you'd expect software giant Microsoft Corp. to have a commanding lead in the race to develop and control many of the most promising technologies and markets of the digital age. And in fact, Chairman Bill Gates and his lieutenants do speak confidently -- and quite seriously -- of a time not far off when Microsoft software will be anywhere that software can reside, from wristwatches to dishwashers to interactive TV networks and programs.

Such heady talk is far removed from the prosaic discussions of word processors and spreadsheets, the personal computer products that have made Microsoft a very wealthy company. The company's profits hover at about 25 percent of revenues, which for the year ending in March were estimated at $4.4 billion. But to evolve beyond personal computers, Microsoft must answer a fundamental question: What is the new business model for success in the consumer world that Microsoft is pursuing so ardently?

Recent profiles of both Bill Gates and Microsoft in various publications have discussed at great length the company's far-reaching product strategy and growing list of alliances (see alliance listing, pp. 4-5), but none focused on how Microsoft might reach its lofty goal of ubiquity while also making money. The company won't talk about it either. But sources both inside and outside Microsoft, most of whom requested anonymity, suggest that Microsoft already knows that the model which served it so well in the personal computer industry doesn't work in the consumer market.

No one doubts that the company will pull out all the stops to succeed in the consumer market. Gates has created a ferociously competitive company that's more savvy about the scope and direction of digital technology than almost any other company in the world. Moreover, it has little choice but to push forward. Corporate computing isn't exactly a growth industry anymore, and Microsoft itself is responsible for driving down the prices of its cash cow -- application software -- to maintain its market dominance. If the company doesn't move forward, it is only a matter of time before it begins to slide.

HOW MICROSOFT DOMINATES THE PC INDUSTRY

Microsoft's business model for the personal computer industry is relatively straightforward: it consists of licensing the operating system (a critical piece of software that, like the grooves in a record album, allows the computer to "play" application software), selling application software (i.e., spreadsheets, word processors and databases) and selling tools and/or developer support to companies like Lotus Development and Borland International who want to create application software.

Microsoft makes its real money from applications software that it develops in-house. It maintains a staggering hold on the market: applications represented 60 percent of Microsoft's $1.57 billion in pretax profit this year, with operating systems contributing 35 percent and hardware 5 percent. Its enormous success gave Microsoft the leverage to start price wars in the applications business -- especially by offering its applications bundled together in what are called "suites" for a fraction of the standalone price. As a result, competing companies that were once high-flyers are barely profitable in their best quarters.

Operating system licenses -- once Microsoft's central revenue source -- are purchased by PC manufacturers such as Compaq or Dell for a flat fee, usually $30-40 per box. Microsoft's initial entry in the market was the MS-DOS operating system, which has evolved to include the Windows graphical user interface. Although the revenues from operating system licenses are relatively small, the software itself continues to be strategic: the key to Microsoft's success in the PC business is that it has linked its three product areas into an interdependent chain. Once the operating system is dominant, anyone in the chain -- from applications developers to hardware manufacturers -- must march to Microsoft's drum in order to stay connected to its millions of customers (or until, as some rivals fear, the company drives them out of business).

DOES THE MODEL TRANSLATE TO NEW PRODUCTS, SERVICES?

As a business model, this simple chain -- operating systems, applications, tools -- maps pretty well to the CD-ROM market, where the popularity of the new medium is helping drive sales of Windows PCs into the home. Microsoft is investing heavily in consumer-oriented software such as CD-ROM titles and children's creativity applications. Its Microsoft Home division has already introduced 45 titles -- with at least 100 more titles planned for release during the next year. Because of its dominance in the PC business, Microsoft has essentially established Windows as the preferred platform for CD-ROM development; nearly every company developing titles creates them first for Windows because the installed base of Windows PCs in the home has far surpassed that of its closest competitor, the Apple Macintosh.

But the model doesn't work so well in the truly promising (i.e., high volume), transaction-based technologies such as interactive TV or online services. Unlike the PC business, where it has great influence over all the links of a relatively short chain, Microsoft has no inherent influence in the large and existing infrastructures it wants to penetrate (cable networks, telcos and content providers). But it is Microsoft. That's good news to those who feel the urge to connect with a sophisticated software provider, and bad news to those fearful of indentured servitude.

In addition, the PC software model -- which only tags operating systems, tools and applications as revenue generators -- doesn't have a mechanism for generating or collecting the royalties on system usage that interactive media applications are expected to generate. But based on what the company has announced in these areas, what it has revealed to potential customers and its historical approach to taming markets, an even more profitable business model is unfolding for the interactive network market that Microsoft covets.

THE QUEST FOR RECURRING REVENUES

By all indications, Microsoft's consumer business model is focused on moving from standalone products to creating sources of recurring revenue.

One reason for the shift, of course, must be that any business model based on hardware is finite and Microsoft's appetite for markets is apparently infinite.

Last summer's rumored (but never consummated) Cablesoft alliance between Microsoft, Tele-Communications Inc. and Time Warner -- the three companies were said to be cooperating on an interactive TV system design -- was the first public indication that Microsoft was serious about entering the interactive network market. Its first major product initiative was announced in mid-May, when it took the wraps off a key software component for delivering video on demand and electronic shopping.

Still referred to by its code name of Tiger, the innovative media server is based on a version of the Windows NT operating system, called NTAS (NT Advanced Server), that uses standard PCs and cutting-edge ATM (asynchronous transfer mode) networking products to deliver video, audio, animation and information services into the home.

Craig Mundie, general manager of Microsoft's Advanced Consumer Technology (ACT) group, acknowledged that Microsoft would charge licensing fees for both the NTAS and the hardware reference design. But even though Mundie and Nathan Myhrvold, director of Microsoft's Advanced Technology Group, would not admit that a royalty structure based on customer usage was in the works, the truth is that the structure of the cable system -- and where Microsoft software could reside within it -- would make it very difficult for any company to make money simply selling a one-time operating system license for interactive TV hardware.

First, there aren't that many cable TV head-ends in the world where a media server can sit -- less than 12,000 in the U.S. And second, even though potential volume for home settop boxes is enormous, their extremely low margins would preclude Microsoft from charging more than a few cents per box.

MINING THE MOTHER LODE OF TRANSACTIONS

With the profits from media servers and settop boxes thus constrained, logic dictates that Microsoft will be inclined to create a different revenue strategy. While Microsoft won't confirm it, it's clear that the company is trying to win support among its potential customers for a transaction-based model that would bring Microsoft a recurring revenue stream.

Here's one of the radical ideas sources suggest Microsoft is floating with network providers: It is offering both to set up and maintain the interactive services part of the Tiger network -- including processing customer transactions and billing -- in exchange for a percentage of revenue that each customer pays the cable company for interactive services, or a percentage of the price paid for each transaction the customer conducts over the network. It might, for example, collect 5 percent of every $3 paid for a movie on demand, or 5 percent of the $20 per month that a family pays for unlimited videogame play on its ITV network.

Assuming a nearly perfect world, where half of today's 59 million basic cable households will have access to an interactive TV network, and assuming again that they're all willing to pay at least $20 for the pleasure, Microsoft could collect $1.2 billion a year just on transactions without selling a single piece of shrink-wrapped software. Now that's a new business model. Of course, the network would need to be based on Tiger software for this to happen, but according to the Gatesian theory that you should never lose a sale over price, Microsoft might even consider giving away the system software in exchange for a piece of the royalty stream.

A DOUBLE-EDGED SWORD

Judging from early looks at the Tiger system, Microsoft seems to have done a great job and the technology alone is a major contender in the server wars -- right up there with competitive offerings from Silicon Graphics, Oracle, Hewlett-Packard and the rest of the media-server pack. This presents the classic double-edged sword for Microsoft's potential customers in the cable community.

On the positive side, Tiger is based on Windows NT, an existing operating system, which brings Microsoft in the door with the momentum of a very large and successful development community behind it. That means lots of good tools are in place to build interactive applications, and the likelihood of getting programming quickly onto a network -- thus the potential for success -- is vastly increased.

In fact, some believe that Microsoft's purchase of Softimage, the Canadian computer graphics tools company, is directly related to this effort. Like Microsoft Office, which is a collection of productivity applications, Softimage contains something called "Digital Studio" -- a collection of advanced multimedia and digital video (read: interactive TV) production tools for high-powered UNIX workstations. Peter Neupert of Microsoft's Advanced Consumer Technologies group says Microsoft will migrate Softimage tools to the PC desktop within five years. (See Cash Flow, p.22)

The flip side, of course, is that once such a system is in place, with all the benefits outlined above, Microsoft will control the technical direction of the network, or at least the interactive part of it. Some fear that Microsoft will then also take over the interactive media programming/applications market, as it did in PC software, because its internal content producers will have early and better access to new versions of Tiger's Windows NT system software than will outside developers. (Microsoft consistently insists an internal Chinese wall between the operating system and applications divisions does not allow this to happen; developers consistently make claims to the contrary.)

It's an intriguing idea to let Microsoft manage the messy details of setting up and maintaining a complex and sophisticated interactive media network. But some cable operators wonder if letting Microsoft manage certain transactions is worth, say, 5 percent of their company's revenue stream for interactive products. They say they'd prefer to pay a flat fee for the technology, and zero percent of revenues.

In addition, despite the fact that some providers believe Microsoft's technology would beat the competition under any circumstances, others are concerned that such a scenario is so comprehensive that it locks out competition and precludes an assessment of a fair market price for the value added.

One content provider even called it a "Trojan horse operating system" because not only could Tiger create special advantages for Microsoft's own developers, but it also provides revenue to Microsoft all the way up and down its existing product offerings of operating systems, tools and developer support.

ONLINE SERVICES, ENERGY MANAGEMENT AND SMART PHONES TOO

Interactive TV is not the only consumer area where Microsoft intends to create recurring revenue streams. Its planned online service and energy management trials via cable have the potential of being at least equally lucrative, if not more so.

The company won't talk about its online initiative -- code-named Marvel -- but plenty has been speculated about it. Again, it demonstrates Microsoft's march toward the Holy Grail of moving its products from a per-unit sales model to one that collects an ongoing revenue stream.

Marvel will be built into a new release of Windows, code-named Chicago, which may come to market in early 1995. Unlike America Online and CompuServe, which require users to install a separate piece of software to access their services, an icon for Marvel will appear automatically on any Chicago user's PC screen. Marvel users will pay Microsoft monthly for a base level of service, an hourly charge after the basic charge and transaction fees for email, downloading files and customer service. Content providers collect only subscription fees, while Microsoft keeps all connect charges.

Given the high level of enthusiasm in the cable industry for connecting to the Internet via cable, Microsoft is also likely to offer fee-based access to Internet "sockets," or built-in network connections, via Marvel.

TIGER AT THE HUB OF "MICROSOFT EVERYWHERE"

It may help to think of Marvel in terms of its relationship to the Tiger media server. If Tiger is the information hub of consumer services, Tiger's settop box is the first spoke in the wheel, and Marvel-within-the-PC is the second.

Third would be the energy management system -- designed by Microsoft, TCI and Pacific Gas and Electric -- that allows homeowners to set up a command and control center for home appliances such as dishwashers using a Microsoft-equipped hardware module and a Tiger-equipped cable connection. Certainly Microsoft and TCI could siphon off a few dollars per customer per month on such a system -- which in an unusual turn of events for an interactive service, might actually save consumers some money.

Telephones, too, could be connected to the Tiger server so that consumers could be charged a few cents per each sophisticated "smart phone" transaction enabled by, say, smart phones equipped with Microsoft software (now being built by Telechips for a trial with US West). Microsoft's Mundie has already announced that Tiger is being designed to handle video messaging, videophones or videoconferencing -- clearly a recurring-revenue service similar to telephony services today.

With its operating system attached to virtually every wire and device in the home, the end result is Microsoft Everywhere, collecting a little bit of money from every activity consumers engage in over the network.

This ambitious strategy is quite a departure for Microsoft, which has been content to collect its money in more traditional ways. But even though Microsoft has a viable technology for the market, it is clearly feeling its way on the business side just like everyone else.

One Microsoft insider says bluntly of today's business models that "they're all broken." That's why virtually all the interactive TV deals that have been made to date -- with Microsoft or its competitors -- are very hang-loose, option-based, non-exclusive letters of intent; if anybody gets spooked about anything, including the intentions of their partners, they can back out of the deal, no questions asked.

So even though fear, uncertainty and doubt usually work in favor of the incumbent, today it seems that Microsoft's far-reaching strategy might work against it. No one wants to take the risk of being married to one system so early in the game, especially if the system has the potential of locking out future choices by controlling the technical direction of global networks. As one cable provider said, "You have to be careful -- you're buying into an annuity that will never end."

DO YOU WANT IT SOON, OR OPEN?

Of course, the question that network providers have to honestly ask themselves is, "Do you want it soon, or do you want it open?" The adoption of any new media has often required one company to control enough of the pieces to drive it into the mass market. A good example was the film business, which in its early days was owned in its entirety by Hollywood studios. The studio owned everything from the film to the theaters it was shown in. After a few decades, the government decided that the system was unfair to independent producers and forced studios to divest their holdings in theaters. But by then, the business of cinema was well established.

Thus the nascent interactive media network scenario today is not a happy one for entrepreneurs or even civil libertarians, depending on your level of paranoia. But the alternative requires a great deal more patience and willingness to cooperate than anyone caught in the digital vortex seems willing to expend. Many organizations, including some of the people studying National Information Infrastructure (NII) initiatives for the U.S. government, would prefer to see a group of interested companies sit down and hash out a set of "open platform" standards for connecting to and providing programming for an interactive broadband network, rather than allow a single company or even two or three to control access. Though this is a terrific idea in theory, it's likely that any such attempt will be seriously compromised by powerful interests along the way.

SERIOUS CHALLENGES AHEAD

The bottom line for Microsoft is that it will have to face down some very serious issues before it can succeed in getting any critical mass for Tiger and its other royalty-based product offerings.

First, it will have to prove that it can be a good corporate partner. The company's reputation as an uncompromising negotiator, to put it nicely, is a very real detriment now that it must work with others to make a system happen. Second, the company will have to be a bit more realistic and forthcoming about whether all these great new products will work. Despite terrific descriptions and demonstrations, virtually none of its most promising technologies are complete enough to judge how they will perform.

And last, there is a very real risk that Microsoft has bitten off more than it can chew. A quick browse through the last few months of computer trade magazines reveals that even Microsoft's corporate customers, tolerant in the extreme, are confused by the sheer amount and complexity of work being done on operating systems alone. Even with 15,000 employees and an R&D budget of $600 million, there's a limit to how much even Microsoft can take on.

With these caveats in mind, one has to question if Microsoft can really deliver Tiger in time for its mid-1995 trials, not to mention start a whole new business as a service provider. Nonetheless, this is hardly a time for complacency for Microsoft's established and future rivals. Though the demands it faces may take a toll in the short term, it would be a grave error to underestimate either the talent or the tenacity of Microsoft.