Information Age: News, analysis & insight for IT & business leaders

Ever since the dawn of the computer industry, one company has towered over rivals. IBM, with annual sales now approaching $100 billion, has dwarfed all comers - today, it is over twice the size of Dell, almost three times larger than Microsoft, nearly four times larger than EDS, 15 times bigger than EMC.

But three years ago, Carly Fiorina decided that she could forge a worthy challenger to Big Blue by smashing together two very different companies: Hewlett-Packard, the flagging industry legend she was running from its headquarters in preppy Palo Alto, and hard-driving Compaq, the Houston-based systems company that had swallowed Digital Equipment and Tandem in earlier years. With its bold acquisition, HP threw down the gauntlet to IBM, saying it would fight it on almost every front - in systems, in storage, in systems software, in services.

Eight reporting quarters on from the May 2002 closing of the deal - and after a sticky first year that left everyone guessing if the combination could possibly work - and HP looks like a company ready to give IBM a run for its money.

Now it can boast an annual run-rate of around $80 million, an industry par growth rate of 12%, a net profit margin that has steadied around 4% and share leadership claimed in some of the IT sector's key market - servers, storage, PCs, and, naturally, its cash cow, printers.

But behind the scenes lies a company with vulnerabilities to match many of its strengths. Upbeat performance in its commodity product divisions masks weakpoints in areas where enterprise customer are demanding demonstrable value-add - and those enterprise issues are a world apart from HP's mass market interests in digital cameras, printers, PCs and industry standard servers.

How well it manages to strengthen some of that critical soft underbelly - and how well customers accept its version of how computing will evolve into a flexible, on-demand service - will determine whether HP is the true rival of IBM or will, in coming years, continue to remain in its shadaow.

Glass half full

What is certain for now though is that HP is a transformed entity from the company that that went into the merger. Wall Street analysts such as Toni Sacconaghi of Bernstein Research believe that sentiment surrounding HP has shifted from "glass half-empty to a glass half-full".

Other analysts such as Ali Irani at CIBC are emphasising how, since mid-2003, HP has built a track record of solid financial performance that has helped allay long-standing concerns about "uneven execution".

And there are major aspects of the business that are very different from the HP of old. Take the example of IT services.

For most of the 1990s, as demand for services such as outsourcing and web consulting was catching fire, HP was proclaiming that it had little ambition in services other than the traditional break-fix maintenance of its own products, preferring to pass any wider consulting, outsourcing and systems integration business to partners such as EDS and Cap Gemini. When in late 2000 it realised that enterprise customers were invariably demanding 'solutions' that involved services as well as systems, and saw how services contacts were often pulling through systems sales at rivals such as IBM and Compaq, it panicked and tried to create an instant high-end services businesses through acquisition - an abortive $18 billion bid for the IT consulting wing of PricewaterhouseCoopers (a deal, in hindsight, that could have killed the company had it gone through).

But when that collapsed due to investor misgivings, HP embarked on a rapid build up of its internal services skill base - almost from scratch - before in 2002, finally establishing credibility and scale in that market by absorbing (through the Compaq deal) the well-established services business that had evolved at Digital Equipment.

The end result is undeniably impressive. The $12 billion services business is currently growing at 15% after a run of high profile customer wins particularly in managed services and outsourcing.

Another stellar area for the company is software - in particular, systems management software. After years of viewing its OpenView product line as niche, the company has taken the tools for managing and optimising systems infrastructure centre stage. OpenView products are critical to the deliver of the HP Utility Data Center (UDC), the on-demand computing platform behind the company's vision of how IT can be an effective, efficient enabler of the 'Adaptive Enterprise'.

Irrespective of how well that nascent strategy evolves, OpenView (and its telecoms software services stablemate OpenCall) sit at the centre of a software business that grew by 23% in its most recent quarter and is heading towards $900 million for fiscal 2004. At this stage, it may be a loss-making division, HP concedes, but that is simply a factor of the investment (both internal and through smallscale acquisitions) required to support its pivotal role in the Adaptive Enterprise strategy in areas such as systems virtualisation and provisioning.

"We believe the fundamental piece of the software stack is the systems management piece," says Peter Blackmore, head of HP's Customer Solutions Group. "The thing that challenges most organisations is not the operating system or the database or other parts, but the cost of managing their infrastructure and linking that to their business processes. Without that you can't get to the notion of IT delivered as a utility, or create an agile business."

But while services and software are clearly areas where the company has executed well, there are issues elsewhere that have some analysts calling for some radical action.

Multiple fronts

"What does HP stand for?" Steven Milunovich, an analyst with Merrill Lynch, asked in a recent research note. "Printers, computers, invention? We consistently hear the computer customers say they're not sure what HP's strategy is. Some customers remark that HP is fighting wars on too many [unrelated] fronts."

HP itself has recognised that it is still too closely identified with its products and not well enough know for its ability to provide answers to business problems and opportunities. That is evident in the recent regrouping of its business units to bring the previously discrete systems, software and services units under the same sales and management structure, the Customer Solutions Group.

But that does not dilute the fact that in some of HP's core product areas there are areas of concern.

Fiorina is often heard to trumpet the message that 'the era of hot box is over', but for all of its re-branding Hewlett-Packard is still best known in enterprise computing circles for its server line up. Indeed, if printers and consumer PCs are excluded from the revenue picture, servers (and their related storage systems) account for the largest proportion of overall revenue. That is not just because of its own server pedigree (HP 9000 and Integrity), but also as a result of inherited lines from Compaq (ProLiant), DEC (Alpha) and Tandem (NonStop).

While that $16 billion-a-year business has been fuelled by a history of innovation - not just at HP but at Digital Equipment and Tandem - the company's declared intention is to move away from some of the core proprietary technologies that, arguably, made those lines successful. It is well down the route of phasing out Alpha and PA-RISC chips in favour of a line - not unlike that of many rival server vendors - that exclusively uses Intel and Intel-compatible processors. At the same time, the intention is to consolidate its operating system activities (that historically have spanned HP-UX, Windows, VMS, MPE and NonStop) around Windows , UX and Linux. The aim is to leverage the lower costs that flow from the use of industry standard building blocks.

But that raises the question: will that approach provide it with enough differentiation, enough scope for innovation, to sustain it as a force in enterprise servers longterm?

Observations by some analysts suggest that with the rationalisation of its server lines it has already started to lose ground - and, indeed, is being squeezed between two formidable forces.

Almost as soon as the merger was closed HP had to cede the number one slot in servers (at least in terms of revenue) to IBM - and it has lost further ground since.

In the most recently available period, for the first quarter of 2004, figures released by research firm Gartner show IBM topping the worldwide server market with $3.6 billion in revenues, a 16.7% rise from $3.1 billion a year earlier. In contract, second-placed HP's server revenues of $3.1 billion were up just 6.2%. That leaves it well ahead of Dell and Sun - whose revenues both stood at $1.2 billion - but aware that Dell is closing the gap fast with a growth rate of 25%.

The upshot is that IBM's share of the market in revenue terms increased from 26,4 to 29.7%, while HP's $3.10 billion gave it 26.9% of the market, down from 27.9% a year earlier. Dell and Sun took 9.8% each. What that hides, though, is the fact that while sales of Intel boxes are growing fast, revenues from high-end systems based on proprietary systems are stagnant.

"With Unix and NonStop revenues are flat, even though we are shipping more [of those] boxes," says Blackmore.

Moreover, HP is able to point to the fact that it shipped more server units than anyone else in the last quarter, as a result of solid sales of low- and mid-end Intel boxes. Analysts suggest HP finds itself in an uncomfortable position. "HP [is] caught in a squeeze," argues Milunovich of Merrill Lynch, with IBM exerting pressure at the high end and Dell winning at the low end.

The company's competitive position has clearly benefited from the economies of scale that flowed from the merger, he says. "Merging with Compaq was a necessary though perhaps not sufficient move to make computers competitive." He concludes that the company would be losing significant sums of money, rather than being modestly profitable in computing, if it had not merged with Compaq.

But as loyalty to older lines peaks, HP needs to come up with a convincing reason to buy from it rather than rivals - hence the advanced thinking on utility computing and solution selling.

Tough at the top

Storage is another area where HP has been knocked off the winners rostrum. In the market for storage systems during the first quarter, EMC grabbed the No. 1 spot back from Hewlett-Packard, fueled by a 26% increase in revenues and giving it a 20% share, according to IDC, up from 17% a year ago. In contrast, HP's revenue share fell to 18% from 19% as sales dropping 0.1% to $630 billion.

Again, HP can point to some better numbers. When all storage - including that shipped internally within PCs and servers - is counted, HP remains at the leader.

The story is considerable more upbeat in PCs. With $35 billion in annual sales, the company has shown it can take on even the mighty business model of Dell and win. Since mid-2003, its revenue growth rate (17% in the company's latest quarter) has surpassed Dell's, while maintaining profitability - albeit wafer thin.

The real profit engine lies elsewhere. Printers and other imaging products still account for the largest part of the company's revenues (30%), and with an operating margin of 16%, the lion's share of company-wide profits. In fact, in 2003 that $1 billion in operating profit contributed 66% of the company's overall profits.

That has led to suggests that HP is artificially bolstering weak or even unviable business units - it has even prompted calls for the company to be split in two.

Most notably Merrill Lynch's Milunovich in early June called on HP to hive off its printer business as a separate company in order to maximise shareholder value. He sees the lucrative operation as being dragged down by the computer hardware division, arguing that the computer unit is now healthy enough to stand on its own.

A separation over the next few years makes the most sense, he said in a research note.

"HP is the Campbell Soup of printing - it owns the category so its main challenge is not share but growing the category," he wrote. "Unlike IBM, HP has distinct businesses that could perform better as stand-alone entities. We think breaking up HP makes financial as well as business sense." However, he does concedes that at this stage, HP management shows no interest in any break up."

The management has other ideas of how to raise its valuation - a valuation that is still trailing where it stood when the Compaq deal was first announced in September 2001. As a tactical measure in late May, for example, they announced the company would use $2 billion of its cash to purchase its own shares in the open market, thereby propping up the stock's value - that is on top of the $256 million share repurchasing authorised three months earlier.

But Fiorina maintains HP is firing on all cylinders and that investors should no longer be jittery about HP's ability to execute consistently across all of its business.

At HP's annual meeting with Wall Street analysts in Silicon Valley in early June, she was feeling bullish about the company's future.

Four years into the CEO's job and more than two years after the $19 billion acquisition of Compaq, HP is in a strong position to continue taking market share from competitors, she said. Revenue growth may not continue at the current rate of 12% but it will be in the high single digit percentage range for years to come. And she bragged that in this fiscal year alone HP's revenues will grow organically by about the size of storage giant EMC - $7 billion.

That puts her in reach of the real goal. "We cannot say absolutely every step has been perfect, but we can say with confidence we are where we intended to be," Fiorina said. "We're the clear challenger to IBM."