CNET también está disponible en español.

Ir a español

Don't show this again

Tech Industry

Taking a second look at HP-Compaq

Analyst Ashok Kumar says it will take a deft touch to pull off a successful merger, but if Carly Fiorina flips the right switches, the potential combo may surprise naysayers.

    Hewlett-Packard's announcement to purchase beleaguered Compaq Computer has been met with a resounding vote of no confidence from the investment community.

    The deal has run into a wall of skepticism because the intrinsic integration and execution risks have been heightened significantly by the checkered execution histories of both these companies--not to mention the significant revenue leakage as vendors like Dell Computer capitalize on the FUD (fear, uncertainty and doubt) factor.

    However, the combination of the two companies does create an incredibly broad product line that in some ways surpasses even IBM's. Compaq's fault-tolerant systems, product lines acquired through its Tandem acquisition, provide the extreme reliability required in key markets. With technology from its Convex purchase, HP offers some of the most powerful scientific computers in the world. Both companies also offer a range of traditional servers from a single processor to dozens of processors.

    Merging product lines
    Compaq's recent decision to abandon its proprietary Alpha microprocessors helped pave the way for this merger. HP was already in the process of phasing out its proprietary PA-RISC microprocessors, and both companies are converging on Intel's Itanium as their server processor of choice. Thus, after some transition period, the combined company can focus on a single server hardware platform.

    The move from proprietary platforms to industry-standard Intel servers is putting pressure on margins at both companies. HP recently terminated an internal Itanium chipset, transferring the design and the design team to Intel, because it couldn't afford the effort. On the other hand, both companies have significant server hardware design expertise and have emphasized hardware differentiation in the past. The merger will allow them to spread the investment in Itanium hardware across a broader range of systems with a larger revenue stream, enabling continued differentiation despite lower system margins.

    The software story is more complex. Compaq's software situation has been a mess due to its acquisitions of Digital and Tandem, each of which had two proprietary operating systems, one Unix and one non-Unix. HP has its own pair of proprietary operating systems: HP-UX Unix and MPE. In the short term, the merged company can continue to support a variety of operating systems on a common Itanium platform.

    Compaq had already planned to shift most of its Itanium systems to the "Monterey" unified Unix developed by IBM and SCO. The merger will put pressure on HP, the only major holdout from the Monterey effort, to merge HP-UX into Monterey, although this is not likely to happen in the short term. Ultimately, we expect the new company will significantly pare back its support for legacy proprietary operating systems. An additional concern will be porting the Tandem fault-tolerant OS technology to the next-generation fault-tolerant Itanium systems.

    Taking on IBM and Sun
    Compared with IBM, the new company will have a much larger share of the Itanium server market, allowing it to deliver better technology at a lower cost. IBM's strategy is to split its efforts across its proprietary PowerPC platforms as well as its Itanium products. We expect Itanium to become the dominant server platform, making IBM's strategy questionable and giving HP-Compaq a leg up.

    The combined company will also be better able to take on Sun Microsystems, which surpassed HP in overall server market share. The merged company will have larger server revenue than Sun--or any other company. Sun's market leadership has given it an economy of scale that helps reduce the impact of its research and development costs. The new HP will have the same economy of scale, plus the benefit of not needing to develop its own microprocessors. (Although HP continues to engage in joint development with Intel on Itanium processors, we expect Intel to acquire HP's processor design team in the next six to 12 months.)

    Furthermore, HP will benefit from the industry momentum behind Itanium, with more than 20 companies developing or selling Itanium systems and many more developing Itanium software. Sun is the only major proponent of its SPARC platform. As the industry shifts from proprietary platforms to Itanium, we expect HP will benefit as the leading supplier of Itanium systems, and Sun will become isolated.

    Challenges ahead
    There are many potential pitfalls in executing this strategy. To realize the projected $2.5 billion in cost savings will require a large number of layoffs, including in the server area. HP has already endured multiple rounds of layoffs under CEO Carly Fiorina, and many have not been handled well, resulting in poor morale. Compaq employees have been trained to view HP in a hostile fashion, and soothing these feelings will not be simple. A major challenge will be retaining key personnel through the massive restructurings that the merger will require.

    A maturing and commoditized PC and enterprise market mandates structural changes to the business model, which will not manifest in the HP-Compaq merger. Focus will shift from revenue growth to revenue maintenance with increased emphasis on cost reduction as the route to profitability. Too many merger candidates lose their revenue momentum as they concentrate on cost synergies. Fluctuations in revenue can quickly outweigh fluctuations in cost savings. For a 1 percent shortfall in revenue, a merger can stay on track to create value if cost savings are 25 percent higher than anticipated (McKinsey, et al). As such, the expected revenue leakage (10 percent to 15 percent) will completely mask the purported cost savings. If they stay apart--still a high probability--both Compaq and HP could at least maintain industry growth rates.

    Bridging customers through the transition will be another challenge. More than a decade ago, HP briefly became the No. 1 workstation vendor by acquiring Apollo; by the time the acquisition was closed, however, many of Apollo's customers had already fled, and HP was never able to wrest the workstation crown from Sun.

    To make sure the new acquisition does not encounter the same fate, HP must quickly assure Compaq's customers that it will continue to support key hardware and software platforms even as it begins to reduce costs by rationalizing the product lines. Does the inexperienced CEO Fiorina have the deft touch required to pull this off? We will soon find out.