Con't: Is BEA Systems the Next Microsoft? Part 2.
Copyright 1999, 2000 and 2001
The Luskin Report
All rights reserved.
Options expressed are those of the authors, and not necessarily of The Luskin Report. Copyrights to individual articles are held by their respective authors.
RULE THE WORLD
Is BEA Systems the Next Microsoft? Part 2.
Paul Philp
Saturday, September 8, 2001
BEA Gorilla Analysis
Is BEA the Gorilla of eBusiness infrastructure software? This is a question worth a detailed analysis since it is a market that will demand a standardized application development platform and environment and the eventual Gorilla will have the opportunity to become "The Next Microsoft". Of course, it should be kept in mind that Microsoft fully intends to be "The Next Microsoft" in this market as well.
1. Architecture
eBusiness is an opportunity which desperately needs a new architecture to remove the complexity of application development, maintenance and operation. The dominant architecture when the World Wide Web began to gain traction was client-server. Early Web applications (web sites) were kept entirely separate from the client-server applications. Slowly, as web sites started to present catalogs, inventory status, take orders and record transaction the web islands started to merge with the enterprise systems.
Initially this integration was still trivial and was accomplished with simple technological hacks such as CGI/ASP. As B2B commerce, complex order taking, customer relationship management, supply chain management, marketing automation and other applications were added to the eBusiness application suite, the simple technology hacks were no longer sufficient and allowing full blown integration of web sites, B2B applications and eBusiness with the client-server enterprise applications started to loom as a complicated, expensive disaster. The client-server architecture was never designed for the always-on, externally focused, cross-functional needs of eBusiness.
eBusiness infrastructure software is aimed directly at solving this increasingly complex problem. The hub-and-spoke system in the airline business is an analogy. If every airline needed to connect each city together individually the combinations would get large very quickly. Clearly this is not economically feasible. The hub-and-spoke system, when each city connects to one hub and most flight are city to hub to city flights. The hubs reduced the number of connections needed to run an airline. In much the same way the eBusiness infrastructure serves as a hub connecting different eBusiness applications with the client-server applications and databases.
Without this new architecture eBusiness is as economically infeasible as the airline system is without hubs. The number of connections that would need to be made could easily be in the thousands and new applications would take years to write.
BEA has taken an extremely open, standards based approach to this new architecture. First, by choosing the open J2EE standard as the starting place for its eBusiness infrastructure platform. J2EE is an open standard developed by Sun Microsystems, IBM, BEA and others. J2EE simplifies application development and decreases the need for programming and programmer training by creating standardized, reusable modular components. J2EE is actually a complex set of standards – comprising 13 different standards at last count. BEA has committed to each of the standards and has typically been the first to market as each new standard has been released or updated.
BEA has extended the open J2EE standards in a number of proprietary ways. The two most important are application integrations /connections and eCommerce components. Although the J2EE standard is open, the way vendors use the standard to integrate with other applications, such as inventory, customer maintenance, logistics, is completely propriety. BEA has undertaken an aggressive campaign to integrate existing applications into the WebLogic server infrastructure. The second important proprietary part of the BEA proprietary architecture is its implementation of reusable components for ecommerce. These include Portal Server for portals, shopping carts, personalization, auctions and many others. BEA intends on growing these proprietary components over time.
It is clear that BEA and the eBusiness infrastructure market passes the architecture test for a new Gorilla. There is an absolute need for the new architecture. The BEA architecture is both open and proprietary. One final note on architecture, this is a market that will only support but one. Software developers will not be able to support multiple eBusiness infrastructures as they develop new eBusiness applications. This is indeed a winner take all market.
2. Network Effects
The next necessary condition for the emergence of a new Gorilla are network effects. Network effects now favor BEA so strongly that is difficult to imagine a competitor being able to catch them. What are these network effects?
Over a year ago BEA launched an aggressive partnering program to recruit software developers and system integrators to the WebLogic server platform. These partners represent are a key prize in this Gorilla battle since their blessing is necessary for BEA to achieve its goal of becoming the de facto industry standard platform. These partners will either recommend BEA to their customers or just automatically include BEA in the sale.
BEA has more than twice as many customers as it’s nearest competitor, IBM. According to Giga Research BEA is also growing its market penetration faster than IBM. As BEA builds its installed base of customers it becomes attractive to the software developers and system integrators. These companies are reluctant to commit to a platform until they are sure they are going with a winner with a large market. As more developers sign up with BEA they bring the BEA products to their customer base and the installed base gets larger. These are the increasing returns caused by the network effects. The value in the BEA network increases substantially with every new partner and customer.
BEA currently has trained 6,000 consultants to implement Web Logic based solutions. The switching costs for these consultants are quite high and once they learn WebLogic and its application integrations and development tools it will be too expensive for that consultant to switch to IBM, Oracle, Sun or Microsoft.
BEA has developers creating 89 repeatable products based of the BEA products and added 130 new software developers in the previous quarter alone and they become locked-in to the BEA platform. Software developers in particular need a standard platform and since BEA has recruited far more software developers than IBM they are reaching the point of critical mass where software developers will work with BEA simply because it appears that it is the standard and no developer wants to pick a platform and then risk having to switch (remember Lotus’s disastrous decision to port 1-2-3 to OS/2 before Windows?).
The network effects are just beginning to kick in and BEA already has over 10,000 customers and 2,000 partners of various types. The value of the BEA network is growing at an increasing rate and has become a virtuous cycle. Recent additions to the BEA network include Dell and Intel. Dell has agreed to ship a trial version of WebLogic server with every server they ship. Intel and BEA have formed a strategic partnership to optimize WebLogic Server for servers based on the new Intel Itanium processor line.
Although the Intel relationship is new and contained, it is hard to understate the importance of this relationship. Intel needs a killer application to drive sales of the new chip and BEA gets privileged access to all the Intel server partners. There are 25 in total and BEA has first shot at a close relationship with 24 (I think IBM will pass). Dell, Compaq, Bull and the others are unlikely to work closely with IBM as they compete directly with IBM on most server sales. The BEA/Intel relationship looks very similar to the Microsoft/Intel relationship and the motivations on both sides are compelling.
The network effects have begun to produce increasing returns and BEA is in the early stages of locking in the market. The strong relationships with developers and integrators pose an awesome barrier for any new company trying to enter the market. Clearly, BEA passes the second Gorilla condition with flying colors.
3. Whole Product
BEA is winning the whole product battle both because it has developed the strongest whole product offering and because its primary competitors are strongly constrained in this regard.
As pointed out in the section above, BEA is beginning to develop a strong software developer community and a strong system integrator community. Two major global system integrators, Accenture and PriceWaterhouseCoopers, have established full blown BEA practices and their competitors cannot afford to be far behind.
When all the pieces needed to build an eBusiness application are laid out, BEA has one or more solid solutions for all the requirements. Some of the key elements of the BEA whole product are:
Hardware integration: BEA has a close relationship with Sun and an emerging relationship with Intel
Security: PentaSafe
Content management: Vignette, OpenMarkets
B2B commerce: Ariba
ECRM: Kana, Broadvision
Application integration: WebMethods, KPMG
The list is long and comprehensive. The key point is that there are no pieces missing for the pragmatic majority to begin adopting the architecture. That is not to say that there are not pieces required for a deeper and broader whole product, it just means that the minimal set is in place to allow pragmatists adoption to begin.
Now, lets look at the challenges that BEA’s competitors face in developing their whole product. The three primary competitors are IBM, Oracle and Sun with IBM the only other competitor with a market share over 10%.
IBM has a excellent application server product with WebSphere as its 20% market share indicates. IBM is late to release their reusable eBusiness component product – San Francisco – and this is limiting the whole product value of WebSphere. However, this is an internal issue for IBM that they will eventually resolve. The primary challenge for IBM is that they compete with all the system integrators with their professional services. As professional services become a larger part of the IBM strategy the large and small system integrators look for opportunities to work in markets where IBM is not directly competing for their clients. System integrators are an essential function in the value chain, especially during the high demand period of a Tornado. IBM has the required professional services to offer customers, however, they will not be asked to give unbiased advice about the selection of eBusiness software. This conflict with the system integrators will be a permanent limitation on the whole product that IBM can offer.
Sun has a market share of about 8% with their iPlanet application server. Sun is the guardian and champion of Java and would seem like a natural leader in the J2EE application server marker. However, Sun faces two obstacles. The first is their role as Java champion. If Sun were ever to reach market leadership their role as custodian of Java would raise the spectre of Microsoft in the hearts and minds of the developer community. Even though Sun has excellent developer tools, it is unlikely that the developer community would support a strong Sun product in the application server market. This cripples their whole product. The second conflict is on the hardware and operating side. Part of the whole product is fully supporting all the hardware platforms that customers require. This includes HP, IBM, Linux and Windows/NT. It is difficult for customers to believe that Sun will deliver and support real ports to such large competitors.
Oracle is entering the market relatively late through the acquisition of Orion. The Oracle product, Oracle 9AS, has about 8% market share. Oracle is a strategic mess these days since it decided to enter the applications market and compete with its customers and large channel partners. Oracle competes with the software developers in applications and the system integrators in professional services. With so much channel conflict it will be impossible for Oracle to assemble the required value chain to compete in this market unless they back off their applications strategy. Larry Ellison is not the back-off type and so Oracle has little chance at a leadership position in this market.
BEA has the required whole product today plus the whole product value is increasing each month. Each BEA competitor has substantial channel conflict to prevent them from developing the necessary broad and deep whole product and value chain. This is a very large advantage for BEA.
4. Tornado Market
If you have made it to this point in the two article set, it must seem pretty clear that BEA is the Gorilla in this market. Unfortunately, life is rarely so easy, as the fourth test for Gorilla status demonstrates for BEA.
On the most recent conference call when BEA announced their financial results for the second quarter they lowered expectations for revenue growth for the next two quarters. BEA set expectations for the current quarter as flat sequential revenue growth and for the next quarter as single digit sequential revenue growth. These numbers are not anywhere near sufficient for there to be a tornado market happening.
A close look at the BEA software license revenue growth over the previous 12 months reveals some evidence that a tornado market did actually begin and has just stalled due to the collapse in IT spending. Add to this observation the fact that several market research firms claim that the application server category has been growing faster than 100% for about 18 months. A case can be made that this tornado blew just long enough to anoint BEA as the Gorilla and the rapid uptake of software developers and system integrators is further evidence. However, the case for BEA being a Gorilla having gone through at least one tornado is tenuous and borderline at the best. Given the recent tendency to throw the Gorilla title around liberally, I urge caution today. It is a close call but I think it is closer to the spirit of the Gorilla Game to acknowledge that there has not been a tornado in this market yet.
However, what can be said is that the pragmatist herd is moving. They are starting development projects and small scale deployments. BEA has significant penetration in the Fortune 500 and Global 2000. The pragmatists have decided to adopt the eBusiness infrastructure architecture and they appear to prefer BEA today. If the economy were to pick up tomorrow, it seems likely that a tornado would immediately launch in this market since all obstacles have been removed and now the herd will just want to get the transition over with. If this happened, it seems that BEA has significant advantages as the network effects accelerate, and BEA would be the likely choice. However, technology markets are unpredictable and it is possible for the tornado to falter.
BEA: The Next Microsoft or The Next Novell?
BEA has the first real chance at becoming the next Microsoft since Novell and Lotus called off their merger in the early 1990s. Although the current downturn has added some uncertainty to the market, it is likely that BEA will establish WebLogic as the operating system for eBusiness. At this point, they will have become the next Novell having earned the right to take on the champ. If this happens there will eventually be a clash between BEA and Microsoft. This will be the hardest fought technology competition since Microsoft killed Netscape. This time the outcome is not certain and a draw is possible.
One way or the other, all technology investors must keep their eyes on this market as the future of enterprise technology is at stake. |
