IBM's software group: can the elephant dance? - IBM Watch - Company Business and Marketing - Column

Software Magazine, August, 1996 by Sam Albert

People are wondering whether IBM's Software Group will be able to shake its reputation as the lumbering giant that not only flubbed desktop operating systems in the battle of OS/2 vs. Windows, but flunked openness and was late to client/server computing as well. Will Big Blue ever find a way to balance its fealty to legacy customers and big iron while developing next-generation client/server software that by definition must cut the cord with the past? Can it realistically hope to keep pace with a thousand tiny, yet nimble, competitors, each of which seems bent on developing a best-in-class product that delivers a point solution with no peer? In short, can the elephant learn to dance?

You might be surprised to learn that more and more Big Blue watchers are coming to the conclusion that, yes, the elephant may not only be able to dance, it might actually be ready to waltz, jitterbug and swing. Still, IBM and its dance masters, Chairman Lou Gerstner and Senior Vice President and Software Group Executive John M. Thompson, had better watch their step. The ballroom they're dancing in is a busy and dangerous place where more than one large competitor has fallen into the orchestra pit while the band played on.

The software industry has never been more confusing, competitive or fast-moving. Software customers, as well as vendors, are rushing toward advanced network computing and the Internet. The old host-centric model, even with its reliability and scalability, can't compete with the combination of low-cost desktops coupled to high-performance servers. IBM, like any software developer, must respond to the needs of individual users and departmental workgroups as they embrace this new paradigm.

During the past several years, client/server computing has enjoyed rapid adoption. But while users can start with a small investment and add applications quickly, they are growing dissatisfied with this model. Client/server systems don't scale up easily and are not robust enough for mission-critical applications. In addition, maintenance and user-training costs for client/server systems are extremely high compared to the cost of supporting a user of host-centric systems.

Meanwhile, the rise of network computing has brought additional complexity and opportunity: Islands of computing within enterprises are getting connected and integrated; enterprises are reaching out to connect with suppliers and customers to increase efficiency and add value; and new applications are being rapidly developed to leverage the network and produce instant advantages for customers, many of whom need new solutions to serve their own competitive markets.

Thanks to its long experience, IBM still leads all other vendors at the enterprise server level of this multitiered computing model. IBM can deliver network connectivity, data repositories, transaction processing, enterprise systems management and software development tools. If IBM can quickly link enterprise servers to Internet and intranet systems -- and let the world know about it in simple terms -- it can win in this space.

However, at the PC client level, Microsoft has hegemony over operating systems, as well as a bevy of winning desktop applications. Still, the emergence of the World Wide Web and products such as Netscape Navigator could move storage and data management functions from the client to the network and pose a threat to Microsoft's dominance. Lotus Notes and Sun's Java programming language, which promise a future where thin clients run disposable applets in a still-newer computing paradigm, would enlarge the threat because they have the potential to change the very fundamentals of desktop application development and use.

Vendors are also fighting for supremacy in the middle tier of the client/server systems space. The Web integrates the PC client with the network, bypassing many middle-tier client/ server products. Major players in this segment include Oracle, Microsoft, Novell and, more recently, Lotus, as well as all the Unix vendors. This battle of architectures is being fought at the operating system level to determine the winners in application development tools and database software. Ultimately, the battle is to gain mindshare with software vendors who must bet their businesses on the platforms, architectures and tools that will deliver tomorrow's winning applications.

In answer to these forces, IBM has adopted a software strategy in which the following objectives will receive full play:

* Establish leadership in network computing;

* Grow distributed middleware (on all platforms, not just IBM's);

* Lead the industry in multiplatform standards;

* Provide integrated, easy-to-use solutions;

* Lead the industry in network application solutions;

* Simplify doing business with IBM.

This is a tall order and requires a substantial melding of IBM's product offerings. At the very least, IBM must recognize its strengths and weaknesses and change its product mix to grow significantly faster than the industry in distributed application development tools, databases and systems management software. Since time is of the essence, don't be surprised if this strategy leads to more headline-making deals like the ones that brought Lotus and Tivoli into the fold.