InformationWeek: The Business Value of Technology

October 30, 2000

Printer ready

Still At The Top?

continued...page 2 of 2

By Aaron Ricadela

More on Microsoft:

sidebar:New Focus On Developers: Q&A With Sanjay Parthasarathy sidebar:X Marks The Box: Microsoft Woos Game Developers Microsoft Solidifies Its .Net Plans (9/28/00) The future of windows (9/4/00) TechWeb: Microsoft Plans MSN Blitz For Late October (10/4/00)

Send Us Your Feedback

Competitors criticize the proprietary nature of Microsoft's strategy. ".Net is being handed down on stone tablets by the voice of God," says Simon Phipps, Sun's chief software evangelist. Sun's Forté for Java environment isn't the most widely used Java tool, Phipps says; products from IBM, Inprise, and Symantec are all more popular. Phipps says that's evidence that Sun provides more choices.

Many of the questions raised about .Net will affect IT departments as well as developers, perhaps more so. After about five years of relative stability in Visual Basic, Visual Studio.Net introduces fundamental changes to the language that will require developers to convert their existing code, says Dan Barclay, CEO of Barclay Software Inc., an Orange, Texas, supplier of banking applications. Though VisualStudio.Net theoretically makes it easier to write server-side apps that run in a Web browser, the necessary code conversion could slow down development and introduce bugs. "I've got a corporate asset in my code, and I'm storing it in Microsoft [Visual] Basic," Barclay says. "I have to wonder if that's a safe place for it now."

Competitors aren't standing still. Oracle VP Rene Bonavie says the company is adding about 3,000 developers a day to its Oracle Technology Network online community, currently at more than 1 million members. Oracle modeled its 2-year-old independent software vendors program on Microsoft's tactic of giving away servers and tools. "You had to buy $10,000 to $15,000 worth of software before you could prototype on Oracle, and people weren't buying," he says.

Microsoft may no longer own the franchise on easy-to-use development tools--a key selling point to Windows developers. IBM's VisualAge for Java, for example, contains libraries of objects, features that eliminate common errors, and links to online help.

In another example of Microsoft's slowing momentum, big applications vendors such as PeopleSoft Inc. and Siebel Systems Inc. have chosen IBM's DB2 Universal Database as their primary database development platform--not Microsoft's SQL Server. "Microsoft couldn't stand still. It had to change," says James Governor, an analyst at Illuminata. "There are new component-based technologies, and Microsoft is going to have to re-educate its community."

And expand it. Part of Parthasarathy's plan involves forming relationships with technology startups that can supply compelling pieces to applications hosted by Microsoft and application service providers. Among the startups on Parthasarathy's radar screen: Xdrive Inc., which provides online storage services, as well as GuruNet Corp. and RUsure.com Ltd., contextual search engines that combine client software with server-side processing.

"Small and midsize developers have a lot of the new, fresh thinking," he says. Aligning itself with emerging technologies, such as peer-to-peer computing, and applying those technologies to business software could also help Microsoft deflect the Java question. "Most cool apps today, such as Napster, don't use Java," Parthasarathy says. "They use a rich client"--i.e., a big, fat PC.

Cultivating relationships with small software vendors is vital to Microsoft's success, says John Sculley, CEO of Apple Computer from 1983 to 1993 and now a partner at Sculley Brothers LLC, a New York venture-capital firm that funds technology startups. Microsoft's size and self-sufficiency make it hard for small companies to figure out who's in charge of projects, and decisions from Redmond often come slowly. "That's been a blind spot," Sculley says. "Microsoft will never be able to make the transformation from a product company into a network-service business without having a large number of trusted relationships with small, entrepreneurial companies."

Microsoft is so eager to see its Web services vision succeed, it's enlisting help from some unlikely sources. Earlier this month, Microsoft bought $135 million worth of Corel Corp. stock, giving it a 24.6% stake in its desktop-applications competitor. Corel plans to use Microsoft's upcoming development tools to transform its desktop apps into services that live on the Web as applets, says Derek Burney, Corel's new CEO.

Moving forward will depend, at least in part, on Microsoft's ability to keep faith with its developers. The vendor's product predilections tend toward IT infrastructure: databases, operating systems, and programming languages. Independent software vendors that build on top of these products to deliver more powerful functionality to customers stand a chance of doing well. But those that compete directly with Microsoft quickly find themselves at a disadvantage, even scrambling to redefine their businesses. And some have had enough.

"The lower-level it gets, the more important you can bet Microsoft is going to consider that technology," says Frank Slootman, VP of software products at Inprise, once a leading vendor of development tools and database software for Windows. Inprise, which is changing its name back to its old moniker, Borland, has since refocused on cross-platform, Java-based tools. Three years ago, about 80% of Inprise's revenue came from its Delphi and C++ Windows development environments; this year, they'll contribute about 40% of sales. "Playing a role in that space is definitely hazardous to your health," Slootman says. So is being slow to market--it was Borland's procrastination with a Windows version of DBase, the market-leading DOS database it bought from Ashton-Tate in 1991, that sunk the product.

Nevertheless, Microsoft's tendency to muscle in on new territory represents a significant challenge as it tries to convince its developer community that Windows-centric technologies will be relevant on the Internet. Over the years, Microsoft has repeatedly taken software that thrived in independent markets--OLAP services, Web browsers, development tools, PC utilities, message-oriented middleware, a transaction-processing engine, and mainframe access software, to name a few--and made that software's functionality part of Windows, SQL Server, or other platform products. "It's a classic--Microsoft decides that something on the market becomes a feature of its product," says Mik Gilpin, an analyst with Giga Information Group.

That modus operandi has landed Microsoft in hot water more than once. The most obvious example is the evidence presented during the Justice Department's antitrust case against the company that it gave away Internet Explorer in order to kill Netscape's business, after Netscape declined Microsoft's offer to divide the browser market. But there are other examples.

In a less widely reported case, a federal judge in September ordered Microsoft to pay $1 million in damages to tiny software developer Bristol Technology Inc. Bristol, on track for about $10 million in sales this year, makes software called Wind/U for porting Windows applications to Unix. Beginning in 1994, Microsoft licensed to Bristol the source code for Windows NT--which the company needed to build its products--in order to gain mindshare among developers. By using Wind/U, programmers could write business apps that ran on the new Windows system and be assured that they could port them to run on Unix workstations and servers. But once Windows NT became established, Bristol says, Microsoft raised licensing fees and denied it adequate technical support in order to squelch competition from Unix. A U.S. District judge agreed, ruling that Microsoft violated Connecticut's fair-trade law by engaging in "wanton, reckless," and deceptive business practices.

"We still write for Windows because Microsoft has a clear monopoly on certain segments of the market," Bristol CEO Keith Blackwell says. But in June Bristol shipped eSleuth, a business-intelligence product for transaction analysis that works with IBM's MQSeries messaging middleware. By next year, eSleuth's revenue should surpass Wind/U's, says Blackwell. "Microsoft will allow you to grow until a certain point when it believes the market's been validated," he says. "You can never win--it's the referee for the platform."

Other companies have found ways to adapt when Microsoft encroached on their turf. Business-intelligence software vendor Cognos was already well-established with its PowerCube OLAP server when Microsoft entered the market. Though some of its features overlap with Microsoft's back end, Cognos co-exists with the software giant by selling an OLAP client that works with SQL Server and other multidimensional data containers to deliver sophisticated functions such as advanced visualization and managed OLAP reporting. "Microsoft makes data warehousing more ubiquitous," says Rupert Bonham-Carter, manager of Cognos' Microsoft alliance. "It relies on independent software vendors to deliver that high-end functionality."

Can developers rely on Microsoft? Given its history, the company had better do a lot to prove it can. Can IT managers rely on Microsoft's latest technology road map? Executives admit a certain level of anxiety in the industry. "That's one of the challenges right now," says Microsoftıs Mendlen. "Without the bits, people are speculating and going bananas." It's a question Microsoft had better answer not only in terms of programming languages, but for the future of the Windows platform in the Internet environment.

Illustration by Anthony Freda
Photo of Phipps by Phil Ryan
Photo of Burney by James Elliot

return to page 1

Back to This Week's Issue
Send Us Your Feedback
Top of the Page