Clabby Analytics, my research firm, creates:

• Advisory reports (for instance, what to consider when buying blade technologies);
• Opinion/counter opinion reports (for instance, I have often challenged the wisdom of buying Itanium-based servers); and,
• Case studies (I go on site and talk to real IT executives and users – and report my findings).

When reading my reports and opinions, I strongly encourage my readers to see out counter opinions – and then balance what I say to the opinions of other respected IT research industry analysts. I believe that by weighing what I say, and what others say, IT executives can reach their own opinions on where truth can be found.

In this opinion, I express my reasons for taking such a dim view of Hewlett-Packard (HP) over the past several years. These reasons fit into three categories:

1. HP’s positioning in the IT industry as a "technology leader" ;
2. HP’s product offerings; and,
3. HP product deficiencies.

I consider all of these topics in the following sections.

I have two issues with this HP statement. First, I strongly believe that HP does not have a "complete" technology product portfolio (for instance, it is missing critical middleware technologies – and the company does not build its own high-end storage). Second – I consider HP’s supercomputing strategy and product set to be broken (so why HP is highlighting its involvement in supercomputers is a mystery to me).

Allow me to elaborate on both points…

On September 15th, 2002, HP announced its 'HP Discontinuance Plan' relevant to HP middleware product offerings. The company announced the discontinuance of HP Application Server (HP-AS), HP Application Server Resilient Edition, HP Web Services Platform, HP Core Services Framework, HP Total-e-Server, HP Process Manager and the Changengine family of products.

The crux of this announcement is as follows:

What this means is that in 2002, HP exited the infrastructure middleware business. And yet the company claims that no other company offers as complete a technology portfolio as it does. I note that IBM and Sun both have extensive infrastructure stacks and middleware products – so clearly HP’s claim is in error.

For those not familiar with the words "infrastructure stack", allow me to explain. When building information systems, systems designers/builders choose hardware, an operating environment, a middleware environment, a management environment, and applications and databases. Each of these environments build on the other – hence they are referred to as a "stack". Middleware is a key component of any stack – its how programs can talk to other programs and how data can be share with disparate data. And failure to own this important element of a stack puts HP at the development whim of its middleware suppliers.

Having worked for a hardware maker earlier in my career, I know that aligning partner strategies and priorities is difficult. (In fact, we had to pay our business partners to port their products to our product line, to implement the changes that we wanted, and to offer support for our products). By taking itself completely out of the middleware business, HP has lost the ability to control its own infrastructure destiny, both up-and-down its product stack. This has, in turn, relegated HP to the role of being a systems integrator that needs to wait for its partners to deliver the functionality it may want in the timeframe the partner wants to deliver it. HP is at the whim of its partners for pricing, for development priorities, and for initiative funding.

I consider HP’s server strategy and product line to be broken in two places: 1) the use of Itanium-based processors; and, 2) the failure of Itanium-based servers in the supercomputing space. Allow me to explain…

• Itanium was at least five years late (I could argue even ten years late) to market; and,
• This processor did not meet market expectations in terms of performance (so today, for instance, IBM’s POWER-based RISC processor significantly outperforms HP with its Itanium-based processors in dozens of industry standard benchmark tests).

When researching HP, the picture that emerges from HP’s own description of itself is that it is a maker of consumer technologies (cameras, personal computers, and other devices); a maker of x86 computer systems (personal computers and servers); a maker of management software; and a provider of IT professional services:

• In the consumer business, HP now competes with a wide variety of consumer product makers;
• In the systems business, HP now competes directly with Dell and Lenovo directly – and on occasion with IBM and Sun in blades/x86 engagements. Given its scale-up systems situation, however, I no longer consider HP as a rival Sun and IBM in enterprise class systems (despite the existence of its Itanium-based Superdome); and,
• HP’s acquisition of EDS makes HP a logical competitor of IBM as well as a competitor of its partners (such as Accenture, et al). It is my belief, however, that HP has jeopardized its professional services channel by buying EDS – so I’m not convinced that buying EDS was a really good decision for HP).

From a positioning perspective, what this means is that HP is now aligned to compete head-to-head in lower margin, commodity-based consumer and computing businesses. Gone are the days when HP was a strong, scale-up rival of IBM and Sun in the computer systems business.  And gone are the days when HP can argue that no other company offers as complete a technology product profile as HP...