Looking tired, HP CEO Carly Fiorina was still all business and polish as she faced some of her most discriminating enterprise customers at Gartner Symposium/ITxpo 2001 in Orlando, Florida.
When it comes to walking the walk and talking the talk of an IT executive, Fiorina had no peers, never once fumbling for words while speaking passionately and comprehensively about HP and Compaq's various businesses and platforms. It was as though a part of her had been attending every single workgroup meeting the two companies held over the last six months. That sort of attention to detail is impressive. But will it be enough to navigate the risky route of the proposed merger?
Fiorina fielded a wide range of questions about HP's future from Gartner vice presidents Paul McGuckin and Eric Rocco who specialize in Unix/Windows servers and IT services, respectively. It was especially appropriate that this pair was chosen to interrogate the HP CEO, given that Fiorina's message was very much about what HP will be doing in the server, data center, and professional services spaces.
According to Fiorina, once the HP/Compaq merger is complete, the combined company will hit the ground running. Part of her reasoning is that both companies shared similar karma, something she said was apparent even before merger talks started.
Both had targeted Intel's Itanium as a strategic platform to move many of its existing customers to. This is a relatively easy proposition for customers already on the Windows platform, but both companies also have enterprise customers on other platforms and migrating them won't be so easy. For example, Compaq has customers running Tru64 Unix and OpenVMS on the Alpha architecture and, now that Alpha has been discontinued, the company's plan was to migrate them to Tru64 or Windows running on Itanium.
Likewise for HP, says Gartner's McGuckin. "The company is preparing to move all of its customers currently on proprietary architectures such as MPE and HP-UX [running on PA-RISC] to HP-UX or Windows." Again, these OSes would run on Itanium.
So, after the companies saw some commonalities and began talking merger, the two started to discuss what a merger might actually look like, even before talking to the bankers. In the days since, the two companies have been feverishly working on a comprehensive set of business and technology plans that simplify the combined roadmap. But those plans could cause migration headaches for customers.
According to Gartner's McGuckin, "All roads lead to HP-UX or Windows XP on Itanium." In other words, existing Tru64 and OpenVMS customers are now looking at HP-UX (instead of Tru64) or Windows XP as their only options. The lack of specifics about how this migration will be handled is the source of consternation for Symposium attendees such as University of Miami CIO M. Lewis Temares who has a significant investment in OpenVMS. Temares wanted to know what tools would be available for migration and who was going to show his staff how to use those tools, but he got no answers.
Remembering some of HP's recent announcements in the Linux arena, I asked McGuckin about the companies' focus on just Unix and Windows. He felt that recent moves by HP in the area of secure Linux would be consumed by the combined company's obsession with Microsoft, which made sense to me. "The relationship with Microsoft will compromise HP's Linux position out of deference to Microsoft," says McGuckin. "Basically, Fiorina is saying that [HP] is going to be the power house Intel PC and server provider and biggest Microsoft partner. On the flip-side, Microsoft has no other obvious choice for taking their products into enterprise data centers." IBM, according to McGuckin and Rocco, has only half its heart in its relationship with Microsoft, and Unisys, Dell, and Sun aren't particularly strong alternatives.
Echoing some of Fiorina's statements regarding how Dell chose to resell storage from EMC rather than developing its own, McGuckin and Rocco said, "Dell will be there, but the company is passive on R&D, and needs to deliver better functionality. Unisys is not a major player anymore and Sun does as little as possible to satisfy customers from the IT services perspective; an important part of any vendor's enterprise strategy."
That pretty much leaves two players -- HP and IBM -- both with the ability to go global, both with very different approaches to their enterprise customers. When talk turned to IBM, Fiorina was fully prepared to rattle her saber, and rattle it she did. Fiorina characterized IBM as an out-of-touch one-trick pony, saying "We are providing a real alternative to the proprietary, integrated 'let-me-give-it-all to IBM to make it all work.' "
Sorry Carly. While IBM, and its Global Services division does take the "give it all to us" approach with some of its customers, this by no means is its only approach, nor is the company a proprietary monolith. I'm not buying your explanation as a differentiator. In fact, I'm not buying the whole spiel, even though your gift of gab almost had me.
At times, Fiorina's own message contradicts itself. On one hand, she talks about the company's emphasis on "unifying architectures" (a reference to Itanium and Windows according to Gartner), and on the other, she touts HP's dedication to open, non-proprietary standards necessary to run a complex heterogeneous environment.
The truth? IBM's WebSphere is no more proprietary than Microsoft's .Net. In fact, many would argue it is more open because of its emphasis on Java. On the Internet side, both support the basic Web services protocols such as XML, SOAP, WSDL, and UDDI. But on the application server side, Microsoft's technology is based on the proprietary Windows, and IBM's is based on J2EE. Fiorina might argue that HP offers the same J2EE capabilities via its Bluestone products running on HP-UX but, then again, with two different approaches to Web services (.Net and J2EE), that's hardly a "unified architecture" plan.
Gartner's McGuckin agrees, saying "what she is saying is code for Intel running Microsoft and this is rhetoric that I would use if I were her. Solaris running on SPARC, or AIX running on IBM's PowerPC are no less open than HP-UX running on Itanium. It's all part of [the] poetry that [she] is trying to weave around us."
Thanks, Paul. Rhetoric--that was the word I was looking for.
Furthermore, on the IT services front, where Fiorina's plan calls for partnering with different professional services firms such as Accenture or PwC depending on the customers and their vertical markets (vs. IBM's singular Global Services approach), there is scant evidence that this strategy will actually work. Alliances, notwithstanding Microsoft and Intel's, are always a scary proposition. On the surface, it makes sense. Instead of trying to be all things to all customers like Global Services, partner with specific experts. But here again, Compaq customer and OpenVMS user Temares is skeptical. "If I need a consultant, what makes her think that she can negotiate a better deal with the professional services firms than I can. In my experience, I'm usually able to negotiate a better deal on my own, than when going through a third party."
So there you have it. HP will lead an alliance of technology companies (Microsoft and Intel) and professional services firms (Accenture, PwC, EDS, etc.) into battle against IBM, with Dell and Sun on the periphery getting dragged into various skirmishes.
When I asked Rocco and McGuckin who would win, neither was willing to commit. But McGuckin remains highly skeptical about HP saying, "We believe that HP has significant hurdles to clear before it can deliver."
While not the death knell for HP (it'll take a lot to kill a company that turns $1.5 billion a quarter), the pulse reading here at Symposium regarding Fiorina's overtures isn't promising.