An analysis of HP’s future strategy, post Carly Fiorina

Carly Fiorina’s undoing was her inability to capitalise on the 2002 HP-Compaq merger, seen as her bet-the-company move. HP is on shaky ground at the moment because its product portfolio has become too large and diversified to manage, and lacks organisation-wide synergy. The printing and imaging business account for a disproportionate share of the profits, while its enterprise divisions lag. The options that stand before HP’s board range from organisational restructuring, to a complete split of the company. Which of the many strategies is eventually adopted depends on the identity that HP decides to create for itself.


A few brief facts about the exit of Carleton S. Fiorina, 50, Chairman and CEO of Hewlett Packard, on Tuesday, the 9th of February 2005: Robert Wayman, the CFO of HP, has been appointed the interim CEO, and has made it to HP’s board of directors. Wayman has been at HP for over 30 years now. Patricia Dunn, a board member since 1998, is now the interim non-executive Chairman.

Carly Fiorina contends that disagreements with board over how to execute the company’s strategy led to her ouster. In a statement, she said “While I regret the board and I have differences about how to execute HP’s strategy, I respect their decision.”

Wall Street reacted favorably to the news, with HP’s stock rising $1.36, or 6.8%, to $21.50 in midday trading on February 9. [18]

HP said that it would start looking for a CEO immediately. The top contenders within the company appear to be Vyomesh Joshi, executive VP of the recently-combined printer and PC business, and Ann Livermore, executive VP of the Technology Solutions group.

What did Carly do wrong?

For all the post-mortem analysis that has been done over the exit of Carly Fiorina, one conclusion that everyone seems to have made is that she lacked vision and focus. The focus part is readily seen, given the huge and diverse range of products, services and markets that HP’s been saddled with after the Compaq merger. Further, all agree that it would be the success or failure of the Compaq deal which would eventually decide Carly’s fate. Carly managed the Compaq acquisition brilliantly, but was unable to leverage Compaq’s precence in the PC business into increased sales for the new HP.[15]

HP had entered the high-end computing market only a year before the Compaq merger. Carly claimed the Compaq merger would give the company an edge over Dell in the low-end PC market. This was clearly an attempt to dominate the entire range of the computing equipment market, all at once. Far too ambitious. Also the merger increased HP’s dependence on the high-volume, razor-thin margin PC business, and brought it into direct competition with Dell [14].

So did HP’s merger with Compaq successfully propel it to the number one position in the PC market? Look at these figures[16] for the answer:

2000 Market Share:
Compaq: 13.1%
Dell: 11.5%
HP: 7.8%
IBM: 7.4%
Fujitsu-Siemens: 5%

2004 Market Share:
Dell: 17.9%
HP: 15.8%
IBM: 5.9%

Even before the Compaq deal, the company had tried to acquire PricewaterhouseCoopers Consulting division, but when the price balooned to $18 billion, HP gave up. A couple of years later, IBM successfully acquired the same divison for $3.5 billion. That failed acquisition, followed by the ease with which IBM managed the same thing, did not reflect very well on Carly’s leadership skills. Analysts have often termed her a great manager, but a poor leader.

Now again, HP combined its PC and printer business in 2005 to be able to sell complete “digital imaging solutions”, as its Pavilion range of computers advertises. The PC business is a notoriously low-margin, slow-growth market – PC market growth is slated to fall below 10% this year.[15] Tying the printer business to the PC one does not make good business sense for the former, because the growth opportunities for HP’s printer business are immense, as we shall see later in this article.

Brand positioning and differentiation:

At the time of the acquisition, Compaq (plus DEC and Tandem) was the dominant low-end computer vendor. It positioned its brand as the premium home and business computer. HP, on the other hand, had a reputation of producing the best computing instruments in the market, especially its scientific calculators. The question was how Carly would manage to differentiate between the two brands when brought under a common management. The answer was not very well. HP could not afford to jettison the Compaq brand because it was too valuable. At the same time, having two product lines for both desktop systems and notebooks made marketing much more difficult. The result is that if you look at HP’s website today, there is only a very feeble attempt at brand differentiation. An average consumer’s puzzled by the two brands and wonders how to choose between them:
Notebooks: For Home and Home Office.

HP Pavilion notebook PCs: Productivity and multimedia tools on the go.
Compaq Presario notebook PCs: Desktop performance and integrated wireless features.

Desktops: For Home and Home Office.

HP Pavilion desktop PCs: Perfect for digital imaging needs
Compaq Presario desktop PCs: Advanced home and wireless networking capabilities

The only shred of differentiation is that the Pavilion line has better multimedia/imaging capabilities and the Compaq line’s geared for the power user. But that isn’t great shakes. There is plenty of overlap between the lines. Look at the specs for the Pavilion and the Presarion notebooks and you’ll find some from the former line virtually indistinguishable from some from the latter. Ditto with the desktops. What makes me choose one over the other? And because of all this chaos, there’s no answer to “why would I buy any of these instead of a Dell or an IBM Thinkpad”?

Poor performance from the Enterprise servers and storage division in the third quarter last year (revenue was down 5%, and losses were $208 million)[19] dragged down HP’s performance in the printing, imaging, PC and services business. This was in spite of the same market segment (servers and storage) showing a resurgence that quarter. Fiorina blamed the mess on problems with a new order-processing system in the US. This caused HP to miss some order altogether, and to ship others by air. She also admitted that the usual “acceleration” in server sales towards the end of the year was “muted”. Promising “immediate management changes”, HP fired the head of Enterprise Sales, the EMEA head of the Customer Solutions Group, and the head of Enterprise sales for the Americas[5]. This was seen as many as a knee-jerk reaction, perhaps unnecessary, and unlikely to have any significant impact on what appears to be a problem in execution strategy.

What are HP’s problems?

HP has a wide range of products and services. There are consumer-oriented areas and business-oriented ones. Consequently, HP has a problem of positioning itself, because it has too wide a portfolio. It tries to succeed in enterprise IT, personal computing, peripherals, and emerging technologies. It competes with Dell in the low- and mid-range PC business, against Sony in the consumer electronics segment, and against IBM in the enterprise. All of these three competitors are focussed on their industries. Since Dell does little spending on R&D, has a focussed approach to cutting operating costs, and sells direct to customers, its operating overheads are 10%, as compared to 18% at HP (and 42% at Sun). IBM, at the other end of the scale, has decided it does not want to be in this business at all, and has sold its PC business.

HP can be viewed as being made up of three major units. The printing and imaging business, the PC business (these to were combined in January), and the high-end, enterprise business (Here too, the hardware, software and services businesses were combined to form the Technology Solutions Group).The printing and imaging business is clearly HP’s strength. It accounts for only 30% of HP’s revenue, but generates 75% of its profits. However, because it is dependent on the performance of the rest of HP, it is missing out on growth opportunities. The PC business is large: HP is the world’s second-largest vendor of PCs, behind Dell and ahead of IBM-Lenovo. But Dell’s efficiency in its delivery model is widening the gap between Dell and the rest. Margins in this business are razor-thin, and the industry consensus is that if you’re not purely in the PC business, then stay out of it. Which is precisely what IBM did. As far as the high-end business goes, there are numerous problems there too. They’ve been discussed in a later section.

An interesting point [20] is how HP became a victim of its own “centrist” strategy, the “Switzerland” syndrome. HP tries to diplomatically offer both Windows and Linux servers on the x86 architecture, without resounding success with either. IBM, in contrast, is a staunch supporter of Linux across its product line, the Power-based pSeries, the Intel-based xSeries, even its zSeries mainframes. Sun has injected itself with an adrenaline shot named Solaris 10, and is pushing the Operating Environment with the zeal of a holy war. Sun has come up with all sorts of innovative ideas – such as its pay-per-use grid computing initiative, to which HP has no answer. Finally, both IBM and Sun have assiduously cultivated developers for their product base, and now have a large developer base, which fuels third-party applications and plugins being written on top of their middleware. HP has no developer base to speak of, and no middleware to write applications for.

Therefore, HP’s printing business is the only one which is doing well now, and which has potential for driving HP’s growth. The questions that HP’s new CEO must answer are:
How can HP leverage its strength in printing and imaging?
How does HP define its identity?
How can HP find synergy across its product lines?
What are the future growth areas for HP?
Where is HP losing market share, and what do HP do about it?
Do we keep HP together or do we split it? If so, how?

Breaking up HP into two:

A section of the industry advocates that HP be split up into two companies, with very different markets to target. This is primarily to address HP’s lack of focus, and to end its identity crisis. One, the enterprise business, would be focussed on business customers – the Medium and Large businesses. The printing, imaging and PC business would focus on consumers, small-and-meduim businesses and retail sales. This is IBM’s key differentiating feature: IBM does not focus on consumers, and even sold its PC business while it was still generating profits. In contrast, HP tries to be an enterprise company and a consumer-electronics company at the same time. Patricia Dunn, the interim Chairman, and Bob Wayman, the interim CEO, are, like Fiorina, opposed to a split-up of HP. During Fiorina’s tenure, spinning off the printer business was discussed at least twice, but rejected each time. In case HP is indeed split, the printer business would be headed by Vyomesh Joshi (current executive VP of the combined PC/printer division). The other business (high-end servers, storage, software and services, collectively known as the Technology Solutions Group) is headed by Ann Livermore (executive VP of that division).

Problems with a breakup: The Enterprise business is sick

The PC/printer division might do well/better, with greater autonomy. But the other, high-end high-margin business, is already underperforming (as can be seen from profit-to-revenue ratios). First of all, HP has a very complicated product line[6], which it had begun simplifying after the Compaq merger[7]. Most of HP’s revenues in this market come from selling high-end PA-RISC and Itanium servers with HP-UX installed (although HP also sells Intel x86-based servers running Linux, Windows and OpenVMS).

After end-of-life-ing Compaq’s Tru64 UNIX OS, and abandoning plans to build AdvFS and TruCluster technologies from Tru64 into HP-UX [4] (instead using Veritas’ Storage Foundation Cluster File System and Storage Foundation) last December, HP seems to be giving out signals that further innovation in HP-UX is slowing down. The delivery date for HP-UX 11iv3 has slipped consistently from 2004 to the first half of 2006.

The other problem is the hardware platform. HP-UX runs on the PA-RISC platform and the Itanium. HP has refused to port HP-UX to the x86 architecture [11]. It announced end-of-life for its HP 9000 series running on the PA-RISC. The Itanium (and the Itanium 2) never gained a decent market share in the first place[8], and have been all but superceded by AMD’s Opteron and Intel’s own Xeon EM64T 64-bit processors with 32-bit extensions, which can run both 64-bit and 32-bit applications at native speeds. In addition, HP’s statements on the Itanium do not match its actions. It has announced that “we continue to see Itanium as the future 64-bit microprocessor”[7], implying also that the PA-RISC does not have a future. But late last year, HP dropped its Itanium workstation line, admitting the shift towards 64-bit extension technology [9], and ended its design work on the Itanium[10]. Also, HP sells over 90% of the Itanium servers sold in the market, implying negligible volumes from the rest of the industry.

Clearly, both the OS and the hardware platform are in trouble. HP, therefore, is increasingly being seen as a reseller of other companies’ technologies, rather than investing in its own.

Maximizing returns from the printing and imaging business:

To repeat, the printing and imaging business accounts for only 30% of HP’s revenue, but generates 75% of its profits. If the company is indeed split, this area of the business could grow phenomenally. There is also another strategy that HP could use[2]: Keeping the company together but turning it into an imaging-centric one. Focussing all product and services lines on the imaging business. Here are some of the possibilities:

Buy Eastman Kodak (valued at $9.8 billion) or Xerox ($14.2 billion) and merge the printing business with the new acquisition. This can open up new customers and can help expand the business further. HP can make such an acquisition. It has $14 billion in foreign profits[2] which can qualify for a one-time tax break if this money is invested in the US. HP’s share price in case of breakup would be $27-28 a share[2], out of which the printing business itself is worth $20 appx.

Printing and imaging also has the most opportunities. HP has 40%+ share in the conventional desktop inkjet/laser printer market, but has not yet entered the high-end networked printers market. Today this is dominated by Xerox, Canon and Ricoh. That is why Xerox is a good acquisition for HP. Demand for photo printers and digital cameras is expected to shoot up over coming years. This makes Kodak attractive, as Kodak has strong prescence in the photo printing market as well as in-store kiosk printing/online printing services.

On the enterprise side, HP could use its expertise in high-performance computing and apply it to emerging markets, where again, margins are high. High performance, special-purpose microprocessors from HP could be used in computer simulations for gaming, weather forecasting, bio-informatics and genetics, and medical imaging. These are all areas where the market either is too nascent yet, or is highly fragmented. HP has an opportunity here to become an industry leader by being first off the blocks.

Conglomerate HP:

The alternative to an outright split-up is to build a conglomerate out of HP, like GE. The advantage of a conglomerate is it can take advantage of market fluctuations. When certain segments of the market are doing badly, the management can focus strategy around the other segments, thus minimising overall impact. In HP’s case, [1] when consumer spending is hot, it can gain an overall edge on IBM, which has no consumer business but competes against HP for corporate contracts; if corporate zooms, HP can then move ahead of rival Sony in consumer electronics. This appears to be a less popular sentiment than either a spin-off of the printing and imaging business, and a split of HP.


We’ve been through a range of issues in the analysis above: Where and why Carly Fiorina failed for HP, why HP needs a change in strategy, and various strategy options for the company. The right strategy is one that strikes a balance between the need for HP to redefine itself radically, and yet maintains the strength of a unified HP. That one appears to be restructuring HP completely around its printing and imaging business. That way, HP can target both the consumer and the enterprise, position itself attractively as the ultimate in everything print and image, and keep itself from a difficult and risk-ridden split. What HP ultimately should do is a tough question Carly’s successor will have to answer.


[1] Fiorina steps down at HP.

[2] Choosing a new HP chief and mulling strategy

[3] HP must open source Tru64 goodies – users

[4] HP laughs off Tru64 promises, welcomes Veritas

[5] New round of HP firings to cost $200m

[6] Hewlett-Packard Server Snapshot

[7] HP Product Roadmaps

[8] Only 3,500 Intel Itanium Servers shipped in 2002

[9] HP gives Itanium the boot

[10] HP sends Itanium designers to Intel

[11] Livermore: No need for HP-UX on x86

[12] Does HP need a new course?

[13] HP combines printer, PC units

[14] HP-Compaq merger: Worth the wait?

[15] Analysis: HP PC Unit’s Future Unsettled

[16] HP Analysis: Carly Fiorina Facing “A Skeptical Business Press”

[17] Fiorina’s fuzzy vision

[18] Where Fiorina Went Wrong

[19] Storage, servers bruise HP earnings
[20] Fiorina falls victim to HP’s Switzerland syndrome

bout the author:
Rahul Gaitonde works as a software developer for IBM’s India Software Labs at Pune, India, chiefly in the areas of filesystems, distributed computing, storage area networks and network security. His present areas of interest are the Linux Kernel, Linux on the Desktop and collaboration software. He also writes frequently on a variety of issues, mostly technology-related ones. houses information about him and his work – along with occasional glimpses of his non-work life. He has 3 articles published on OSNews previously: “The KDE 3.2 Beta 2 User Review”, “Giving XFce4 a Spin” and “IceWM – The Cool Window Manager”.

If you would like to see your thoughts or experiences with technology published, please consider writing an article for OSNews.


  1. 2005-02-24 8:57 pm
  2. 2005-02-24 9:20 pm
  3. 2005-02-24 9:48 pm
  4. 2005-02-24 10:15 pm
  5. 2005-02-24 10:46 pm
  6. 2005-02-24 11:03 pm
  7. 2005-02-24 11:12 pm
  8. 2005-02-24 11:20 pm
  9. 2005-02-24 11:21 pm
  10. 2005-02-24 11:26 pm
  11. 2005-02-25 12:06 am
  12. 2005-02-25 12:12 am
  13. 2005-02-25 12:32 am
  14. 2005-02-25 1:42 am
  15. 2005-02-25 2:40 am
  16. 2005-02-25 4:00 am
  17. 2005-02-25 9:22 am
  18. 2005-02-25 1:56 pm
  19. 2005-02-25 2:07 pm
  20. 2005-02-25 6:49 pm
  21. 2005-02-25 7:18 pm
  22. 2005-02-25 7:29 pm
  23. 2005-02-25 8:22 pm
  24. 2005-03-04 8:47 pm