I would never claim to have the software expertise of Paul. My strength is in strategy and seeing the big picture. As such I just did a three part series on BEAS and whether or not it is a Gorilla or not.
From a strategic perspective Paul is absolutely correct, IBM cannot and will not be the Gorilla in this marketplace. The Gorilla forms because of a dire need for a marketplace to become efficient. MSFT is the simplest example, software developers needed a common platform to develop on in order to maximize their addressable market and minimize their costs. Consumers needed a common platform in order to maximize their utility. INTC arose for the same reasons. In the end this common platform enabled the commoditization of the product, and completed the whole product for mass distribution.
IBM's strategy of profiting from services and selling main frames and tossing WebSphere in as a bundle is in complete contravention to this commoditization strategy. IF WebSphere becomes such a complete product that it practically makes distributed computing for most of the world plug 'n play, IBM loses its business model. IBM's model requires continuing complexity.
BEAS on the other hand is shooting for the MSFT model of commoditization and plug 'n play. INTC is pursuing the same strategy with its IA-64 chip. It wants to take the one large and expensive Sun or IBM server, with its single point of failure, and replace them with cheap clusters of INTC based servers, servers sold by DELL or Compaq on-line.
BEAS is INTC's partner in this endeavor for two reasons (1) INTC is competing with SUN and IBM, so not going to use iPlanet or WebSphere, (2) BEAS has the best clustering technology in the market. Clustering is a software program that allows many servers to work as one, and enables you to add or subtract servers, as needed in a plug 'n play environment. This clustering technology allows a firm to say buy 3 cheap INTC servers, and cluster them so as to avoid a single point of failure, and also to allow easier and greater scalability (all at less cost) then say having to buy one large SUN or IBM server, with its relatively high cost, single point of failure, and much more difficult cost to scale if you want to add additional servers. As such, BEAS clustering technology is the key enabling technology to enable IBM to pursue the commoditization of server strategy to ebusiness.
If history is our guide (and I think here it is appropriate), as Moore's law continues, INTC's commodity servers, like INTC's PCs, will over the rest of the decade come to dominate the mass market, pushing SUN and IBM servers back up to the top of the food chain, into a high-end niche. And INTC will once again come to dominate the largest and most profitable mass market. BEAS WebLogic will be the operating system for these clustered ebusiness servers.
IBM is counting on BEAS and INTC to be incapable of commoditizing ebusiness, and bringing it much closer to the plug 'n play environment that we enjoy on the PC today (this plug 'n play environment is not perfect on the PC, and it won't be perfect in ebusiness either, but I think you get the point). If, however, BEAS and INTC can bring this sort of plug 'n play functionality, IBM's army of consultants will be less and less necessary, and why would the vast majority of the marketplace want to buy expensive single point of failure main frames when cheap, plug 'n play clustered servers, with plug 'n play application software, built upon WebLogic will do.
In a nutshell that appears to be the battle in the marketplace today. And it seems to be a battle that is down to IBM and BEAS with no other contenders. IBM is currently enjoying a lot of momentum; but even with the momentum is still #2 and is not necessarily indicative of the long-range result in the market. Like in the PC market, BEAS & INTC will not be able to produce the plug 'n play, commoditized product overnight. It will take years. But if they succeed, it will blow this marketplace wide open, and IBM will find itself in the high-end niche, with BEAS and INTC dominating the mass market.
To put it into perspective, Windows '95 is only 7 or 8 years old. Prior to this the world was living with DOS, and even DOS only became a mass market product 20 years ago or so. The application server market is pretty much in the early stages of DOS in terms of developing the whole product. I think it will develop faster than the PC product did, but it will still take many years.
Of course, IBM might try to change its strategy long-term as well. Given its reliance on consultants and main frame sales this seems doubtful, but that could toss a wrench in the mix as well. So despite IBM's momentum, I think it is BEAS's and INTC's market to lose, but they have to execute and there is no guarantee that the ebusiness environment can truly be commoditized in the way INTC and BEAS envision. What we do know, is that if they can, it will be a far better product than the service and main frame intense market vision IBM is pursuing.
Thus, only BEAS and INTC are taking the truly Gorilla strategy in this market. And the fact that they both much execute well to get there indicates that neither is a Gorilla in this marketplace at this time.
Tinker |
