IBM announced that it has reached an agreement to acquire for $129 million CrossWorlds Software, a Burlingame, California-based maker of business integration software. CrossWorlds produces software designed to integrate and automate customer service, manual entry processes, sales and order processing. IBM said it intends to incorporate CrossWorlds products into its WebSphere middleware line of products. IBM and CrossWorlds had a significant relationship prior to the announced acquisition. CrossWorlds’ products have been integrated to a large degree with IBM’s WebSphere offerings already, as CrossWorlds has had an OEM agreement for IBM’s MQSeries and provides connectivity with IBM’s WebSphere products. This relationship between the two companies spans four years. The purchase of the 350-employee CrossWorlds is an all cash deal and is expected to close in the first quarter of 2002.
In the acquisition-happy days of the Internet bubble this story would not have made very much news at all. A $129 million deal would have barely appeared on most people’s radar screens, which were calibrated to detect much more expensive and grandiose couplings that left people gasping for air as they wondered just how the high valuations of non-revenue generating companies could go. But here and now, with the dust of the speculative frenzy slowly settling, giving technology vendors and buyers a chance to see what the future really looks like, this purchase makes a great deal of sense for IBM. Middleware is very geeky stuff that is all but invisible to most users of information technology. It has long been the domain of IT professionals who have been told their task is “connecting everything to everything.” IBM has been getting very good traction with its WebSphere offerings, and adding CrossWorlds’ technology will only provide the company a better grip for dealing with the changing demands of IT services. CrossWorlds’ offerings provide the very core of line of business needs: customer records and relationship management, customer order and invoice automation, real-time product pricing automation and improved order processing. None of this stuff is glamorous, but if it doesn’t happen, or at least doesn’t happen efficiently, customers, sales and revenues are easily frittered away.
In the larger scheme of things, we see middleware’s ability to comprehensively and coherently bring disparate systems together in a robust and reliable fashion as the emerging dominant task of IT departments in the coming years. As the Service Computing model continues to evolve and take shape within enterprises, vendors offering road-tested solutions to IT departments that must fulfill the mission of providing a stable and reliable infrastructure for their emerging Service Computing environments are going to be very popular, indeed. We also think it is interesting to note that IBM made this rather modest (by earlier standards, at least) acquisition after a significant relationship with CrossWorlds, in essence spending four years of dating. This was no impulse coupling motivated by the need to make some sort of splash in the marketplace. Instead, we see this as part of IBM’s ongoing strategy to position itself as a provider of the nuts and bolts that will bring the value of Service Computing to its customers. With this acquisition, IBM continues to construct a foundation on which emerging Service Computing environments can be successfully built and sustained.
Sun Microsystems has announced the availability of two new entry-level servers, the Sun Fire V880 and the Netra 20. According to Sun, the Solaris-based V880 offers mid-range performance at a lower cost than competitors’ comparable Wintel-based servers do. Additionally, the V880 takes advantage of Sun’s Free Solaris Binary License Program, which allows users to run the Solaris 8 OS on an unlimited number of business and personal computers with eight or fewer CPUs. The Netra 20 is a rack-optimized machine designed for the extreme conditions common in some enterprise, telecommunications and government computing environments, and is the first Netra product to be based on Sun’s latest UltraSPARC III processor. The Sun Fire V880 is priced beginning at $29,995 for a two-processor system with 4GB of memory and six 36GB hard drives. Pricing for a low-end Netra 20, with a single processor, 512MB of memory and a 36GB hard drive begins at $11,495.
As we have discussed previously, while the low-end of the server market may not provide the glamour of high-end servers such as Sun’s recently released StarCat, what they lack in glitz they more than make up in bread and butter revenues. Sun, which has been as badly knocked around by the stumbling economy as any enterprise hardware vendor, is using both the V880 and Netra 20 to plug some noticeable gaps in its sales strategy. The company is positioning the V880 as a direct competitor to increasingly popular Wintel servers such as Dell’s PowerEdge products, which have been gaining ground on Sun in SMB and work group environments. Along with attractive price/performance, the V880 also offers access to Sun’s Free Solaris license program, which the company hopes will entice businesses to abandon their commitment to Windows products in favor of Sun’s Solaris-based products and freeware UNIX goodies. The Netra 20 is essentially an upgrade of the company’s previous entry-level Netra product, but it arrives with Sun coming under increasing pressure from rivals. IBM, in particular, has been making inroads on Sun’s position in the telecom market, as evidenced in recent deals the company announced with Telia and Sonera. In both cases, scores of dedicated Sun and NT servers were replaced with single IBM zSeries machines running hundreds of virtual Linux and NT servers.
While market share advances and retreats are endlessly changeable, the real issues to consider in this announcement are how effectively these two products reflect and advance Sun’s overall strategy. Sun is positioning the V880 as a robust, affordable machine that will release SMBs from the chains of the Wintel combine, but just how realistic is that view? SMBs, especially, depend on Wintel-based products for a wide range of desktop computing and office productivity solutions, easing their migration to higher-end, Intel-based machines from Dell, Compaq, IBM and others. Sun has ploughed a great deal of money and effort into its alternative StarOffice productivity solutions, but their user base remains miniscule compared with the number of business customers using Microsoft Office and other Windows-based tools. While current Sun customers will likely be pleased by the price/performance of the V880 and its practicality for workgroups and remote offices, we seriously doubt that it will inspire a migratory groundswell of users away from Wintel-based solutions. Sun’s upgrade of the Netra to the UltraSPARC III should incrementally improve the system’s overall performance and satisfy current Sun customers. However, the Netra 20 does not address the growing threat that Linux-based solutions pose to Sun’s proprietary UNIX products. Does this mean that Sun is headed directly down the tubes, or that Linux is poised to achieve immediate supremacy in business computing? Not at all. It merely suggests that despite the rapid evolution of enterprise computing, with vendors abandoning their proprietary chipsets and variants of UNIX operating systems for Intel Itanium-based/Linux solutions, Sun continues to rely on its traditional, simplistic assertion of Solaris and SPARC superiority to drive sales and customer loyalty. That may have worked well enough in 1998, but things are considerably different today. We keep wondering, shaking our collective heads, if this is really all there is to Sun’s server market strategy.
General Motors and its subsidiary Hughes Electronics announced on Monday the signing of an agreement with EchoStar whereby Hughes will spin off from GM and merge with EchoStar. The new combined entity would be named EchoStar but use the DirecTV brand for its services and products. EchoStar would have more than 16.7 million subscribers, of which 1.8 million are National Rural Telecommunications Cooperative (NRTC) and affiliates, and 14.9 million are owned and operated by the combined company. The companies noted that at present cable TV operations companies presently control more than 80% of the U.S. pay television market, while the new EchoStar would provide service to about 17% of the market. Multiple stock exchanges and external financing will be needed to complete the merger as will approval from separate and collective classes of GM shareholders, as well regulatory approval from the FCC and qualification from the IRS that the Hughes spin out will be considered a tax free transaction. Assuming all conditions are met, the parties expect to close the transaction in the second half of 2002.
While most will undoubtedly focus on this announcement’s impact on satellite television users, what we find interesting is the potential impact this merger could have on midband and broadband Internet access. Although the numbers pale compared with the 16+ million TV subscribers, DirecTV through its DirecPC product has over 400,000 users of its Internet access services. We believe that the computer-related potential of HDTV and opportunity to address demand for higher speed Internet access throughout the lower forty-eight states, independent of last mile infrastructure issues, caught the eye of EchoStar. Since EchoStar has only recently begun offering these services, merging with Hughes gives the company an instant jumpstart in this potentially lucrative marketplace.
Fixed wireless or satellite Internet offerings are relatively new, but in certain geographies, their deployment is growing rapidly. The consolidation of the competitive satellite carriers should bring increased efficiencies in operations and capacity, and could provide the critical mass needed to effect new converged broadcast and data services. Notwithstanding, one could argue that this would result in a monopoly of these services with no price competition. However, in many markets, competition would exist from cable and DSL services. Since EchoStar services would cost the same to deliver regardless of whether the subscriber is in St. Louis, San Francisco or Tonopah, NV, the company would be hard pressed to justify pricing by geography, and the competition in metropolitan areas should help define the market price for services. Given the current propensity of regulators to approve significant consolidation in the telecom and cable industries, we believe it is only fair that the same standard be applied to the competition. In reality, the more pressing issue for data consumers will be which set top box and related hardware will ultimately be deployed as EchoStar’s and DirecTV’s existing systems are incompatible, and DirecTV-based deployments have a significant lead in satellite data services.
IBM has announced that a number of significant IT vendors have joined in the company’s eLiza Project, designed to offer more sophisticated infrastructure management and security. Among those companies joining IBM are BMC Software, Candle and Nortel Networks. Customers joining this initiative include Danske Bank, Merrill Lynch and Terra Lycos. The eLiza Project is designed to provide a more automated response to business process infrastructure problems as they occur and, in an increasing number of cases, fix those problems automatically. The eLiza Project is part of IBM’s larger “autonomic computing” vision, in which increasingly complex computer networks and infrastructure become self-managing.
When you say the phrase “autonomic computing” today to most CIOs the reaction ranges from rolling eyeballs to belly laughs to being chased out of the office by a very angry person wielding the heaviest object at hand. And that’s just today. As we move forward into an increasingly demanding computing environment that requires much greater stability and reliability, IT departments are going to find their prime directives altered substantially. Instead of determining what applications with what features reside on which individual desktops, IT is going to find more and more of its time taken up with ensuring that the underlying infrastructure of the computing environment is functioning as it should. In this Service Computing environment, line-of-business people will be making more far-reaching decisions as to how each of their desktops are configured, and their assumption of this role is going to drive productivity gains that have to date been largely figments of vendor imagination. LOB control of what applications are available to employees, and how those applications tie directly into what an individual employee does during their 7.5 hours at their desk, will drive the actual computing tasks done by employees further to the core of the day-to -day operations of the enterprise, and thereby create even higher demand for a stable and reliable computing environment.
Meeting that requirement will become IT’s primary task. eLiza – and other offerings – that provide an increased capability for “self-healing” infrastructure environments will become crucial to IT departments as they find the demands of their emerging responsibilities, initially at least, daunting. The ability to provide improved self-healing intelligence within the infrastructure that will have to possess greater-than-phone-system reliability is a requirement of the emerging Service Computing model. As a major service provider, it is essential that IBM develop the infrastructure intelligence that makes IT’s job more manageable, if no for no other reason than to maintain CIO sanity levels. While there is likely never to be a time when the computing infrastructure requires no human oversight, the need for increased intelligence and self-management will grow proportionately with the complexity of that computing environment. Complexity and indispensability are inevitable. eLiza is the beginning of an effort to manage this complexity. It’s either that or skyrocketing CIO felony assault rates.
Digital Island has announced an extension of its 2Way Web Services offerings to provide enterprises managed storage solutions for both distributed and centralized computing. The new managed services were developed through an alliance with EMC, and will be facilitated in part by locating EMC Symmetrix networked hardware and software solutions at Digital Island’s global data centers in order to provide customers Capacity on Demand storage solutions. The services include Fibre-based services, scalable capacity on demand storage for UNIX and Windows-based clients, back-up services for Oracle and Microsoft SQL Server databases, disk-based copies of critical e-business data and remote data protection solutions for mirroring primary data storage to a second geographic location for disaster recovery and fail over. Target customers for these services include large enterprises across vertical markets including financial services, high technology and media and entertainment. No pricing for the services was included in the announcement.
Digital Island is not the first ISP to offer hosted, managed storage services, but we believe several elements of the announcement are worthy of further consideration. While the massive IT build out we have witnessed over the past few years has left many enterprises with more costly computing infrastructure than was probably necessary, the need for acquiring additional data storage capacity remains constant. But given the continued shortage of IT staff in typical multi-site enterprises, effectively operating and managing expanding in-house storage resources has become increasingly difficult. Digital Island’s Capacity on Demand model, which allows clients to add storage in blocks as small as 55GBs, appears to offer a storage outsourcing model that addresses many of these problems. Additionally, the company’s inclusion of geographically separate disaster recovery/fail over services may interest businesses concerned about possible future events linked to the atrocities of September 11.
Beyond these practical issues, we believe Digital Island’s managed storage offers an interesting model for Service Computing, where IT-based solutions will increasingly be offered in much the same way that utility services are today. While the notion of outsourcing business processes is hardly new, business technologies have finally reached a state where robust computing and networking infrastructure can be designed and deployed to support the practical, affordable enterprise-level solutions that will form the core of Service Computing. We are also interested to note that Digital Island will focus its initial solutions at key vertical industry markets. To our way of thinking, the ability of outsourced service providers to develop vertical industry expertise will be key to the success of Service Computing. From that standpoint, its new managed storage solutions suggest that Digital Island has taken a large step in the right direction.
The Sageza Group, Inc.
Veterans Blvd, Suite 500
Redwood City, CA 94063-1743
650·366·0700 fax 650·649·2302
Europe (London) 44·020·7900·2819