9/05/01 Surprising many, Hewlett-Packard and Compaq announced on September 4th that HP would purchase Compaq for $25 billion and merge the computer systems vendors into a single organization, vying closely with IBM for the Number One spot. That assumes, of course, that revenues, in a merger, are additive. IBM is currently a $90 billion company (and profitable); HP and Compaq, both struggling in a tough market, together have $87 billion in revenues. We’d note, however, that combinations of computer companies (or computer companies and other high tech ventures) are rarely additive, or even successful. Later in this article, we’ll look at what’s happened to others, but first let’s look at HP and Compaq. In this article we’ll be referring to the new, combined company as HPC. HP is a well-respected systems vendor, looked at as a worthy (if smaller) competitor to IBM. It competes mainly in the hardware business, for desktops and servers. It falls far behind IBM in the services business and has been trying to come up with an effective strategy to increase its services presence, first trying (unsuccessfully) to buy Price, Waterhouse’s IT consulting business then looking at partnerships and smaller acquisitions. In one area, it is an undisputed leader: HP’s line of PC printers have trounced the competition and given it a strong presence in offices (and homes) everywhere. Compaq can help HP with additional market share in both desktops and servers and substantial additional presence in the service market, via its prior acquisition of Digital Equipment Corporation. Both companies have a strong interest and presence in the storage market. Neither company has a substantial presence in the software market but both provide infrastructure software plus operating systems, drivers, and development tools for their hardware offerings. HP also provides a Web Services architecture, an extension of its earlier eSpeak offering and Compaq has an extensive SAN software offering for its storage hardware. (HP indicated in the HPC announcement that servers, storage, and software would be a $23 billion business for the new company; we’d guess that not more than $3-5 billion of that would be software.) IBM has been moving, in the last few years, towards its goal of making half its revenues in software and services, a goal it achieved in 2000. With 150,000 IGS employees in its services businesses, it fields the largest global services presence; the company has over 300,000 total employees, much slimmer than the more than 410,000 at its height in 1992, but still the most formidable competitive force in the industry. HPC’s work force, in its entirety, will be about 145,000 and this total is likely to decrease somewhat after the merger, as replications in administrative and marketing positions are removed. (Fiorina says this decrease will amount to about 10% or 15,000 employees over the first 24 months of the merged company’s operations.) This makes HPC not only a smaller company, but a different kind of company than IBM, more focused on hardware manufacturing and channel marketing and less able to focus on software development and big ticket sales to very large customers that require more direct customer contact and much more consulting in both the sales and the implementation process. WHERE IS HPC LIKELY TO EXCEL? Hopes here are likely to focus on capturing the number one spot in desktop and server sales. We think HPC will find it difficult to do this considering world market conditions, but possible. In the case of desktop sales, winning the prize may be losing since it’s likely to be unsatisfyingly unprofitable. HPC is estimated to have $15 billion in service revenue, but this is less than half of IBM’s $33.2 billion. On the other hand, more hardware sales can lead to more services sales, on both the maintenance and the consulting fronts. HPC could also consider some software acquisitions, particularly in the infrastructure arena. This could make it more competitive with IBM and Sun and provide an additional source of revenue (and an additional basis for its consulting expertise). This is a great time for buying software companies – prices are depressed and lots of good companies are looking for an exit strategy. Timing is an interesting issue here. It’s likely to take six to nine months for a deal this size to close – a mixed blessing since in this economy, things could be going much better and HPC could start in on an up cycle. On that other hand, it’s harder to be disciplined about making tough changes once the economic engine starts turning again; now it would be easy. Also, the acquisition bargains available now might quickly melt in an up economy, HPC PROBLEMS The first problem, of course, will be getting through the antitrust process. We’d guess that with some negotiation this should be possible. HPC will not be significantly more concentrated in any market except PC desktops and servers than other vendors are today and both of these are markets with other strong competitors. Compaq has no printer business and neither company is significant in the software business. Their services business will be stronger together than separately, but still much smaller than IBM (and only when measured in terms that include significant non-consulting services). The biggest problem HPC is likely to face is the management distraction and loss of productivity caused by the integration process. They need only look to the mess Compaq has made, over the past few years, while trying to integrate their Digital acquisition. Nearly all of the skilled management they acquired is gone. Much of their opportunity to build a services business was squandered by lack of focus and apparent management disagreement on the company’s direction and priorities. We’d assume HP CEO Carly Fiorina has a strong view as to how an HPC integration should proceed, but we’d also note that it’s tough to integrate different cultures (both work styles and lifestyles), to say nothing of different platforms, priorities, and management teams. We’d guess that the number of areas in which HP and Compaq currently have product overlaps will provide lots of room for dispute and negotiation. This includes nearly everything but will probably be particularly noisy in the areas of desktops, servers, handhelds, and storage. It’s important for HPC to come out of the process with a single, well-integrated product line to serves all of its customers well. This won’t be easy – there’s lot to choose from – and the politics could get nasty. HISTORICALLY SPEAKING, THE COMPUTER INDUSTRY RECORD ISN’T GOOD The facts speak for themselves. We’ve had lots of attempts to merge computer empires and very few successes. Generally speaking, such attempts fail because:
DO ANY COMPUTER MERGERS WORK? In fact, some acquisitions/mergers of technology companies do work. We’d observe they have particular traits in common.
We can, of course, give many examples of mergers that fit these rules. Quite a few of Microsoft’s technology acquisitions would fit handily, but especially its acquisition of Vermeer for its Front Page technology and its entrée to the Web as well as Microsoft’s recent acquisition of Great Plains to take it into new markets and new marketing channels. IBM’s acquisition of Lotus nearly fits, except for their decision not to entirely integrate the acquisition, but rather to allow it to remain autonomous for some time, in order to leverage its brand and channel relationships. It is now nearly entirely (six years later) integrated into IBM. We suspect that IBM will move more quickly with its newest acquisition of Informix, which is scheduled to be integrated within a year. Palm’s acquisition of Be and Apple’s acquisition of NeXT (both to gain operating systems and the expertise of the selling company’s famous founders) might also be looked at as “small acquisitions” by bigger companies. Computer Associates, the software company that acquisitions built doesn’t fit into either category. Rather it is a law unto itself. But it has done that by having a crisp (some would say ruthless) model, often buying companies at very favorable (low) prices and then stripping out every non-essential element (administration, marketing, etc., etc.) to add additional products and customers while maintaining CA’s admirable margins. More recently, they have strayed from this model, paying higher prices to buy larger firms and keeping firms intact in an attempt to leverage their channel strategies. This has made it much harder to keep their earlier business model in place. It has also led to much publicized customer dissatisfaction which CA is now reorganizing to better manage. But CA and Microsoft are not systems companies, so perhaps they’re simply outside the model HPC must fit. We are never very optimistic about these big technology mergers. There are so many ways to get it wrong and so few to get it right. Nevertheless, there are a lot of good people in the combined HPC camp (and many more available in a down-cycle high tech industry). If HPC can focus on the goals and have the discipline to insist that petty rivalries and cultural issues be set aside, there’s a chance. We – and you – will just have to wait to see how this one comes out. REFERENCES HP http://www.hp.com/hpinfo/newsroom/hp-compaq/index.htm New York Times http://www.nyt.com/ ClieNT Server News www.linuxgram.com Interactive Week http://eletters1.ziffdavis.com/cgi-bin10/flo?y=hKeR0CFVzS0FUD0Msg0AV
|
