|
INTEGRATION WATCH: BEAting a Path to Nowhere?
By Andrew Binstock
 June 15, 2004 — In late May I attended eWorld, BEA’s annual shindig in which it rallies the troops, announces new products and explains strategy. BEA has always been good at presenting these shows. Numerous technical sessions demonstrated the company’s deep understanding of enterprise applications and running them on Java platforms. No doubt about it—BEA has the technical chops.
What the company apparently lacks, however, is a business strategy that points to growth and shows how the company will hold off IBM on the high end and the growing open-source competitors on the low end. For analysts like me, this was what we hoped to learn at eWorld. BEA customers who might be anticipating future app-server purchases were similarly looking for reasons to believe.
In my opinion, eWorld failed to deliver. The company’s announcements were just a repackaging of known BEA technologies. Let’s look at this more closely. Five years ago, BEA was soaring. It owned the Java app-server market. IBM was far behind it—more challenged by Sun than by BEA. Then what happened?
Within a few years, IBM overtook BEA because of two crucial factors: IBM had a much better entry to its customers (what with selling hardware of all stripes, a market-leading database, and the 800 lb. gorilla in the middleware space), and more important, IBM realized that an app server is not a technical sell. BEA, in contrast, continued to view its product as a technical pure play that had to appeal to developers. So, when IBM started chewing up BEA’s lead, BEA responded with initiatives and technologies that targeted developers rather than its real customers, IT and development managers.
This misperception led to a series of products and releases that have gone nowhere: BEA’s WebLogic Workshop and the dev2dev initiative. WebLogic Workshop, which I wrote about in the Oct. 1, 2002, issue, was originally packaged as a Web services product, not a Java development tool. It has now been expanded and repositioned as “a full-featured Java development environment.” It’s a great product, but who wants to be competing against Eclipse? Note how Borland’s JBuilder and Sun’s NetBeans are faring in that competition.
dev2dev was another attempt to appeal to individual developers. The idea, presumably, is that if developers see how much support they can get from BEA for their projects, they will push their companies to buy its Java server. Most developers know that such a goal has a rather unrealistic aspect to it. Worse yet, dev2dev is nothing close to the real heavyweight of vendor-sponsored Web sites, the much admired developerWorks from…dang!...IBM.
BEA has yet to recognize that WebLogic’s success arose because the product provided an enterprise-scale solution at a time when no one else offered one. It had nothing to do with the desires of individual developers.
Indeed, developers are not a good audience for building large, successful companies. In the past 25 years, only Borland and Rational have had any success pursuing this strategy, and neither one has known great financial success because of it. The problem with selling to technologists is that you can’t really leverage what you’re selling. How do you leverage an app-server sale? Sell services? OK, so you show the customer how to install, configure and run the package. Then what?
Contrast this to IBM’s approach. It goes to a site to solve a specific problem. It sells the site hardware, a Java server, middleware, a database, content management and security tools. It bundles them into a solution and then sells services that use the technologies to tailor a specific deliverable. Then it provides ongoing support for all the IT areas its technology touches, even hosting certain applications, if necessary. Which business model looks more appealing to you, as a development manager?
So, I was hoping that BEA would announce, let us say, that it was now going to start providing IT services for Java shops. It would do integration work with any and all database vendors, and it would tie BEA into your Web site and provide you with all the software and skills you need to do massive transactional work on e-commerce sites. In addition, its new security tool knows how to bolt down your site and make sure nothing gets to the app server that shouldn’t be there. One of these panoramic scenarios would show me that the company understands enterprise-level IT and intends to be a major solutions provider there.
Instead, we heard about Liquid Data, the company’s newest product initiative aimed at service-oriented architecture (SOA). Now, I admit SOA is one of the best architectures ever to come down the pike. The problem is it will appear in IT shops on a purely incremental basis. No one is tearing out his infrastructure to replace it with SOA. Notice the size of all the Web services vendors today (Actional, Cape Clear and the like.) They’re specks. This is BEA’s new target? I expect continued disappointing earnings from BEA until the company understands that solutions rather than technologies are what its customers want.
Andrew Binstock is the principal analyst at Pacific Data Works LLC
|
|
Advertisement
Click here for a complete listing of Integration Watch Columns
Click here to see a complete Column Archive. |
Back to Top
|