Please feel free to forward this newsletter to as many friends as possible. I have a personal goal of having one million readers. I can achieve it with your help.

the Software View: Software market dynamics (Part IV)

Welcome back, gentle reader. Every brand new software market requires a so-called "killer application" or "killer-app", for short, to make it truly successful. The Lotus 1-2-3 spreadsheet was the "killer-app" for the original DOS-based IBM desktop personal computer. The Netscape Navigator Internet browser was the "killer-app" for Windows 95 because it took advantage of pre-emptive multitasking and multithreading to download content from the web. I submit that the Starr Report was the "killer-app" for the Internet. The Starr Report drove millions of international users around the globe to the Internet. It was just a bit racier than Lotus 1-2-3.

For those of you with Internet and World Wide Web access and a Netscape Navigator web browser, please point your browser to:
http://www.softwareview.com/
Scroll down the page and you will spy a link entitled, "Current views web log (fresh news daily! Click reload, clear the browser cache)" The daily news page is also known as a "web log". It is en vogue and the fashion of these days to call it that. Click on the link, click "reload" on your browser or clear your browser cache to ensure that you always receive the freshest, hottest daily news concerning Java, Linux, XML, and the software industry! The link never changes, but I will be updating the HTML file page behind it every day. Please, do take a gander at it every day.

Also, some important news, gentle readers. the Software View is now an Associate Internet World Wide Web site of Amazon.com. I would like to extend my sincere, heartfelt gratitude and thanks for your patronage. I am offering links to books, et cetera that you can purchase from my web site. I would greatly appreciate it if you would purchase software industry books from my web site. Help support my newsletter and web site by purchasing items from Amazon.com from my web site. Here is the URL (Uniform Resource Locator):
Click here

Now, dear readers, on with this week's episode of the Software View!

With this issue, I continue a multi-part exploration of the dynamics of Internet software markets.

James Aley writes, "id Software of Mesquite, Texas, is the company that created Doom, one of the all-time megahit computer games. Id hopes to do it again with Quake and Quake II, ultraviolent shoot-'em-up's that are even bloodier than their predecessor.

It is not just high production quality and Cro-Magnon appeal that's created the mania for the company's games. The egos at id are scoring with a brilliant - if unorthodox - business plan: They spend one or two million dollars developing a game, then give it away. To be more precise, they distribute an abbreviated but fully functioning version of the game free of charge over the Internet. "We want 'em to play till their fingers bleed," says Jay Wilbur, id's "biz guy." If the "game slaves" decide they want the whole shebang, they have to pay for it either at a store or via an 800 number. The Net result is that this puny company has put its games on millions of hard drives around the world - and made $16 million in 1995.

Id Software has mastered the peculiar economics of the software industry.

That the software industry operates under economics of any kind may come as a surprise. At first glance, the industry seems like the demented revenge of a bunch of nerds who failed freshman economics: Give away a million dollars' worth of product? Take public an outfit like Netscape, with $17 million in revenues in the first six months of 1995, and get a market valuation of $2 billion? Are you kidding?

But successful companies like id understand software economics as if it were - as a Quake instruction manual might put it - smashed into their heads with a club. Sure, id starts out by giving away product, an approach made easier and perhaps necessitated by the ubiquity and culture of the Internet. But id and Netscape are just adding an extra twist to basic rules that are ingrained in the strategies of Microsoft, Oracle, and every other company that's turned software into a fortune.

The beauty of the software business - or its curse, depending on who's talking - is that once a company gets a firm foothold in a market, that foothold can morph into a fortress. The economic principle that best describes this phenomenon is "increasing returns." The basic idea, according to one of the leading theorists of increasing returns, W. Brian Arthur of Stanford University and the Santa Fe Institute, is that in some industries there's a natural tendency for whoever is ahead to get further ahead. In other words, winners win big, and losers lose big. Arthur explains that virtually every high-tech industry incorporates aspects of increasing-returns economics - but software is the perfect specimen.

Here's how the game works:

SELL MORE - EARN MUCH, MUCH MORE.

The first rule of software economics is that this business is distinctly of the high-fixed-cost/low-variable-cost variety. Most of the product cost of software is fixed up front in the form of research and development, so per unit costs go way down as volume rises. This is why gross margins in the industry often reach 90%: The more you sell, the more profit you get. And since the production cost of each unit shipped is comparatively small - near zero when the Internet is the distribution channel - even a tiny company like id can manufacture software in unlimited supply.

So far, so what? Lots of industries have high fixed costs, like semiconductors, to name another high-tech genus. But unlike the chip business, software involves little or no raw-material cost and little or no inventory cost. It's just 1's and 0's. And of course, software developers don't need billion-dollar fabs to get started. "Never in the history of the industrialized world has there been a business that has the advantages of scale of the software industry," says Bill Gurley, a director at the DMG Technology Group, a division of Deutsche Morgan Grenfell.

Simple arithmetic makes the reason for this clear, explains Gurley. Say there are two companies selling the same kind of software application for $325 a pop. Each company spends $250 million on research and development and $50 a copy on variable costs like advertising, making the disks or CD-ROMs, printing the manuals, shrink-wrap, etc. But company A sells nine million units while company B sells only one million units. Company A's pretax profits are $2.2 billion, and company B's are $25 million. Company A's return on research and development investment is 890%; company B's is 10%. Note the discrepancy. "You keep investing in technology because you get increasing returns to investment," says Eric Schmidt, Chief Executive Officer at Novell. "This is an industry where it's natural for temporary monopolies to emerge."

This leverage of fixed costs gives the volume leader in a software segment an incredibly powerful weapon: price. To really nail the competition, all the market leader has to do is cut price. In Gurley's example above, say company A trims the price to $250, forcing company B to do the same. The total market expands by 2.3 million units, so now company A and company B sell 11.1 million units and 1.2 million units, respectively. After A's price cut, A is still making $2 billion in profits and a 788% return on research and development investment. Company B, however, loses $10 million, and its return on research and development investment has gone negative.

Now that is market power. Whoever has major market share reaps the rewards and creates a daunting obstacle to challengers. "In the software industry, one player has thermonuclear weapons, and the other guys have bows and arrows," says Philippe Kahn, CEO of Starfish Software, as well as founder and former head of Borland International.

Kahn should know. Once upon a time, Borland tried - and failed - to dominate the markets for database and spreadsheet applications. Borland's defeat was partly due to strategic missteps. It acquired a number of companies in an effort to gain market share, but consumers never gravitated to applications bearing the Borland brand. The more important reason Borland lost is that few understand the power of market share better than the company that ultimately won - Microsoft. The m.o. of this money machine is essentially this: Establish a customer base for a product, then make a killing selling upgrades. Once you've got the market share, the easiest and most reliable way to acquire a stream of recurring revenue is getting customers to buy the latest and greatest version. Last year Microsoft got more than half its retail revenues from upgrades, says Jeffrey Tarter of Softletter, a trade newsletter.

In software, the big usually win bigger. Microsoft spent $860 million last year on research and development. "Every day we get up and ask, 'What should our company be more open-minded to?'" explains the world's foremost software economist, Bill Gates. "In our research and development budgets, you're going to find a lot of things that are just very broad, where we're just trying things out." That buys good technology and stupendous competitive advantage. "If you're in the software industry and do not have dominant market share in your category, you should be nervous," says Gurley. "If you are in the software industry, do not have dominant market share, and Microsoft is entering your category, you should call a board meeting and consider drastic alterations to your business plan."

Even the Internet, where distribution costs are virtually zero, doesn't substantially change this fixed-cost leverage. True, the Net reduces barriers to entry: A tiny company can have an audience of millions without fighting to gain shelf space at CompUSA. But even if you remove all unit marketing, manufacturing, and distribution costs from Gurley's equation, the market share leader still winds up with an enormous gap in return on research and development investment.

THE MORE PEOPLE USE IT, THE MORE VALUABLE IT GETS.

Software is subject to what economists call "network externalities." This essentially means that the more widely used something is, the more people value it. Take Adobe Illustrator. If graphic designers everywhere use Adobe Illustrator, it makes sense for a designer buying his first illustration software package to go with the flow. In effect, says Microsoft group vice president Nathan Myhrvold, "much of the value is delivered beyond the product per se." This distinguishes the software industry from other intellectual-property businesses like Hollywood and publishing. Seeing the latest hit from Disney allows your fifth-grader to maintain her status among classmates. But she can't use the fact that she goes to the same movies as all her friends to make money.

The value of software increases even more if the company that makes it can persuade other parties to create additional products that work with it. One of the sharper tines of id's multipronged strategy is to make its hell and demons-themed games programmable. Players can create whole new scenarios and characters. ("Any character you can think of, fictional or nonfictional, has been used," biz guy Wilbur says with pride.) Customized Quake episodes and Quake levels are everywhere. Some are sold retail by third-party vendors, but the vast majority have been concocted by John Q. Quake-addict and are available free over the Internet. Id doesn't make a dime off the add-ons, but their ubiquity keeps Quake, now 212 years old, from getting stale. By cultivating a steady stream of fresh hells, id has established a self-perpetuating market for its game.

Microsoft's strategy is similar: Create a platform that people can build on, and run with it. "People don't buy Windows just to use Windows," says Myhrvold, "but to take advantage of the different hardware and different software applications. If you're clever, you can create a volunteer army of companies building around your product, and create a buzz that helps people choose your product."

Autodesk is another virtual army raiser. The company pulled in $534 million in 1995 from its line of computer-assisted-design software, most notably AutoCAD. But Autodesk-compatible products and services constitute a subindustry with sales of $2.5 billion a year. There are thousands of add-on products that enhance AutoCAD, and hundreds of independent training centers around the world. "I've got upwards of 10,000 other companies that really want us to be successful because we are why they are successful," says Autodesk CEO Carol Bartz.

ONCE THEY'RE THERE, THEY HAVE A HARD TIME LEAVING.

Once people learn a complex program like AutoCAD, there are powerful incentives to stick with it, as Autodesk well knows. Getting up to speed takes lots of time and money. The cost of switching to a new product may well outweigh the benefit - even if the new product is in some ways superior to the old.

Just as much to Autodesk's advantage is the fact that so many digital drawings are written in the AutoCAD format. Bartz estimates that there are a billion AutoCAD drawings around the world. Each drawing has, on average, ten other files attached to it, like spreadsheets detailing construction costs. Converting to another format may be more trouble than it is worth.

The economics term for creating a customer base that is essentially captive is "lock-in." The first step for software makers looking for lock-in is creating technology good enough to convince people it's worth committing to. "Users don't want to go with the wrong technology and be stranded," says Garth Saloner, a Stanford business school professor who studies how corporations relate to high tech. Saloner calls this the "penguin problem," because it is reminiscent of the behavior of penguins on an ice floe. They all want to dive in and feed, but none wants to be first, lest it become feed to an unseen predator lurking in the water below. Says Saloner: "It's kind of like consumer behavior - do you want to be the only one stuck with the wrong technology?"

Not everyone accepts the idea of lock-in, least of all companies that are skilled at it. To Microsoft in particular the term must sound like fingernails on a blackboard. Concepts like lock-in and increasing returns are used unrelentingly by critics who want to sic antitrust regulators on Microsoft.

Lock-in connotes permanence, that a dominant technology will endure even if a better technology exists. Myhrvold dismisses the idea. "Saying that inferior products can become more popular is fatuous," he says. "You have to take a more customer-oriented view." Myhrvold concedes, though, that when switching costs are high, it takes a dramatic improvement for a new product to dislodge the incumbent. "If 5% of customers think some new technology is better than the old one, that's not going to do it," he says.

Myhrvold's right: Lock-in isn't necessarily forever. For one thing, massive success creates unforeseen burdens. Gurley points out that the owner of a locked-in technology can be just as trapped as the users. For years, Microsoft programmers have had to try to make every new product work with the thousands of DOS-based applications still being used out there. "It became harder and harder for Microsoft to implement new features to Windows and still maintain compatibility with DOS," says Gurley.

What's more, there can always be some brand-new thing that unlocks lock-in, overcomes fixed-cost leverage, and turns the market on its head. A famous example from outside the computer industry is compact disks, which took over the music market despite the fact that music lovers everywhere had entire collections in an incompatible format, vinyl. In the PC spreadsheet market, Lotus 1-2-3, which dominated, was trounced by Microsoft Excel because Lotus was too slow to produce a good Windows version. Now, says Myhrvold, "I love Excel, but if someone comes out with a dramatically better spreadsheet, Excel is in trouble." Sun Microsystems' Java^TM is a pure object-oriented platform conceived from a clean slate. It harkens from the influence of Bill Joy and the world of consumer electronic devices.

THE IMPORTANCE OF BEING EARNEST.

The least "economic" piece of the economics of software is the role played by hype. Victory in the software industry is at least partly determined by which CEO or programmer convinces the most people that his vision is right. In an environment with such rapid change, people yearn for someone who knows - or at least seems to know - what it all means. "There's a whole game of increasing returns that occurs at a psychological level," says John Hagel, a leader of McKinsey's multimedia group. "Confused people are receptive to people who seem to have vision and confidence." A current example of the vision thing is the "$500 network PC." If you believe Oracle CEO Larry Ellison, this is the wave of the future. Ellison's infectious enthusiasm is matched only by Bill Gates' infectious indifference. It's true that Ellison and the companies allied with him first have to deliver such a box, but a lot rides on which man people believe.

Of course, evangelizing often seems like simple deception. ("It's not safe to operate heavy machinery for a few hours after talking to some of these people," says Paul Saffo of the Institute for the Future, a Silicon Valley think tank.) Some tactics are manipulative, such as the use of "vaporware," products companies announce - with no intention of actually releasing - in the hope of taking momentum away from competitors. But Hagel thinks something more substantive and legitimate is going on. "Some people really do have a clear view," he says. "If they're aggressive in communicating that view, and if they're credible and generate that first wave of investment, it becomes a self-fulfilling prophecy." In other words, the path to lock-in starts with good buzz.

These rules of the economics of software are now being adapted to the industry's new world: software that works with the Internet. On page five of Netscape's IPO prospectus, the company declares its intention to make its software the "de facto standard" of the Internet. This may as well be a passage out of a high-tech economics textbook. It's a classic example of one company trying to conquer a market by exploiting the dynamics of increasing returns.

Netscape wants to create a completely new architecture with its own virtual army of developers and applications. By giving away a terrific product - its Navigator Web browser - Netscape made it clear that computer users are ready to embrace the Web. That, in turn, lured thousands of third-party software developers, who now write add-on products, called "plug-in's", designed to work with Navigator.

Netscape's broadside is just the most obvious threat to Microsoft, the desktop computer world's primary setter of de facto standards. Another economically savvy competitor is Sun Microsystems. Its new programming platform, Java, allows propellerheads to write programs that will run on any platform - IBM clones, Macs, workstations, etc. By making Java work on any operating system, Sun hopes to reduce the switching costs of people currently locked in to Windows.

Sun's Java strategy has unfolded spectacularly during the past year. Java now seems to have a reserved slot in the toolkit of every software developer on the planet, even though much of its promise has yet to be fulfilled. Ann Winblad, a prominent Silicon Valley venture capitalist, says: "They've delivered enough so that people believe in it. It's not only a technical feat here. This is a marketing and positioning feat."

Of course, you can't make money by just giving away hot technology. Application developers and end users get Java for free. But third-party developers who want to embed it in their products - like Apple, IBM, Netscape, and Oracle - have to buy a license. For the record, Java has not earned a profit. Not yet, anyway. "But that's the plan," says Bill Joy. "We want ubiquity first, then profitability." He won't give sales numbers or divulge how much Sun has invested in the product, other than to say that "several years" lie between Java and black ink.

A big risk for Sun lies in how Microsoft will respond between now and then. Any challenger threatening Microsoft's hegemony has a nasty fight on its hands, in part because of Bill Gates' legendary willingness to jettison previous plans and change course. When customers say they want something, Gates reacts. In the latest example of that strategic suppleness, Microsoft too has licensed Java and will engineer the software into Windows 98 and Windows NT. "Making sure we are a leader in the Internet is the near-term goal we're really striving for," says Gates. No one doubts that they responded effectively to the threat of the Internet tidal wave. Now the world awaits to see if they will miss the Java train.

The uncertainty of how all of this will shake out is highlighted by a little piece of boilerplate in Netscape's IPO prospectus, in the very same sentence about attaining de facto standardhood: "... there can be no assurance that such goal will be achieved." No joke. Learning the lessons of software economics has never guaranteed success of Microsoftian (or even idian) proportions. There's even less of a guarantee now.

The competition over standard setting and market share intensifies when everyone understands the rules. "When Microsoft is the only one that realizes market share is everything, they win bigtime," says Gurley. Now that the secret is out, the intensity of the drive to set standards on the Internet can push the price of a piece of software to zero. That's the case with Web browsers and numerous other products - what's a little forgone revenue to a company trying to establish a standard? The problem, of course, is that $0.00 is an unsustainable price. Sooner or later, companies have to make sales. Those that don't, lose in an ugly way.

Right now there's no conventional model for translating freeware into profits. Id has taken one approach. The most prominent company to make the transition is Netscape. Giving away its browser led to sales of server software to corporations, which in turn helped spur the development of intranets, which in turn led to the sale of browsers to those same corporations - this time with special features and customer support. Netscape's quarterly revenues have gone from zero to $55 million in 26 months. "We'd like to say it was planned," said Eric Hahn, Netscape's former senior vice president of enterprise technologies. "I think the success has surprised everyone, including ourselves."

In the years ahead, companies like Netscape will have to depend on alliances as never before. In a world where software can be distributed free, upstarts seeking to challenge the status quo need partners to spread their technology as quickly as possible. Those seeking to maintain the status quo will need partners as well to protect their fortress. "It's all about positioning, strategy, and owning market real estate," says Winblad, the venture capitalist. "That means not only building a good product but also building relationships with other companies. It's like the board game Risk. You cannot win alone. You've got to persuade someone else who's playing to be your partner."

Look at the history of Sun Microsystems' Java. Who would have thought that a sleepy hardware manufacturer of workstations and Unix products would create Java? Java has the potential to become the de facto programming platform for the Internet and the standard for cross-platform executable content. Every company of note in the software industry has committed to Java development: Intel, Novell, Oracle, Hewlett-Packard, Netscape, SAP, even Microsoft.

It's hard to believe now, in this age when IPOs, market valuations, and reality seem to travel in concentric orbits, but the notion that software can create value wasn't always obvious. Only seven years ago, when Winblad and her partner John Hummer were trying to launch the first software-only venture-capital fund, they found it hard to win backers. "It took a hundred meetings before we convinced enough investors that software even met the criterion of 'risky venture,'" she recalls.

Until the mid-Eighties, Netscape's Jim Clark was one of the skeptics. "I thought it was a flawed economic proposition, that there's no way you're going to make money on something people can just copy," he says. "It's amazing what watching Bill Gates get wealthy did to my perspective."

To be continued ...

Sincerely,
Mark Kuharich

Join my free e-mail newsletter called the Software View by clicking here or sending mail to thesoftwareview-owner@west-point.org