HP's Path Forward Under ApothekerHP's Path Forward Under Apotheker

Hewlett-Packard's choice of Leo Apotheker as the company's next CEO surprised many, but there are no surprises in the challenges that await him as leader of the world's largest technology company. HP must deepen its presence in software and services or it will face low profits, low growth and a marginalized position in the IT industry.

It's unanimous that the selection of Apotheker, ex-chief and 20-year veteran of SAP, is a confirmation of the importance of enterprise software to HP's future. That point was also underscored by the appointment of Ray Lane, former Oracle president and COO and now a venture capitalist, as nonexecutive chairman of HP's board.

But what's the best path forward in a market in which rivals like Oracle and IBM and hard-to-acquire partners like SAP and Microsoft have many of the biggest and best names in software sewn up? As I've commented previously, just going out and buying software companies, willy nilly, is not the answer.HP's best hope lies in a two-pronged strategy. In the short term, there's no choice but to forge deeper partnerships with the likes of SAP and Microsoft. Creative product and service packages could blunt whatever momentum the so-called "integrated stack" players have managed to generate. For the long term, HP must make strategic investments in growth areas such as cloud computing infrastructure, platforms, applications and related services.

Despite initial surprise at the selection of Apotheker, some note that there just aren't that many global CEO types out there with tech industry experience, and even fewer with software industry experience. (An HP board spokesperson said the company had a short list of six candidates, but that Apotheker was the only one offered the job.)

Compared to the CEO choice HP made nearly five and a half years ago, when it plucked Mark Hurd out of the relative obscurity of NCR (a kiosk and point-of-sale-system vendor), Apotheker is a celebrity.

What Hurd lacked in big-name recognition (at the time) he made up for in relevant experience, having spent 25 years with a computer hardware manufacturer. Apotheker, in contrast, is all about what HP lacks, not what it is today. The ambition is clearly to become a bigger player in enterprise software, but some question whether Apotheker's experience is what HP really needs to get what it's after.

"Consumer innovation is really what's stimulating offerings for the enterprise these days," observes analyst Frank Gillett of Forrester Research. "It's not that enterprise experience isn't relevant, but you need somebody who can blend the two worlds."

Honestly, though, there are even fewer CEO candidates who have straddled the consumer and enterprise technology arenas. And even if HP found one, where could such a wiz apply consumer-Web innovations within HP? By some estimates, software currently accounts for less than 5% of HP's sales, and most of the software the company does have is aimed at IT types who don't demand slick, user-friendly interfaces.

Consumer-savvy innovation is obviously important to HPs PC, printer, and fledgling smart phone businesses. But the company has to really get into the enterprise software game before it can bring Web 2.0 innovation to that market.

Set a Course for Innovation

On the topic of innovation, HP can't count on much of any until it unwinds the tight clamp Hurd put on R&D investment, cut from about 4.0% of net revenue when he joined the company down to about 2.5% in 2009 (by comparison, IBM reinvests about 6% of revenue in R&D).

HP knows this is a sore point. Thus, last week's press release on Apotheker's appointment made the prominent claim that he "transformed R&D and technology platforms" while at SAP.

By most accounts I've read, the engineers and technologists in Waldorf were only too happy to see Apotheker go. Jim Hagemann Snabe, SAP's new co-CEO, is one of their own. And soon after Apotheker was let go, SAP elevated technology chief Vishal Sikka to the board, sending yet another message to customers and employees that technology -- not sales -- would once again drive the company.

Apotheker is not a technologist, let alone a Steve Jobs type. But let's give him the benefit of the doubt for now. By delegating to the right people and reversing HP's R&D budget cuts, he could set a new direction favoring innovation. And those innovations could (eventually) be applied in both the consumer and enterprise markets.

Get Into the Enterprise

So how should HP go about diving more deeply into the enterprise software market? The quick, costly and foolish way would be to buy into the enterprise applications market that Apotheker helped create. Yes, HP could quickly gain big enterprise market share by acquiring SAP, but that would not be the smart long-term move.

"Even IBM has looked at potentially buying SAP… but that's really buying into the past," says Ray Wang of Altimeter Group. "

Much as Apotheker might like the sweet revenge of acquiring and once again leading SAP, HP could get much more bang, buzz and long-term growth for the buck by buying small, up-and-coming software-as-a-service vendors. Salesforce.com is pricey at the moment, with a valuation estimated at about 95 times earnings, but you could put together a balanced portfolio drawing from the likes of SuccessFactors, RightNow, NetSuite, Workday, Adaptive Planning, Host Analytics, PivotLink and others.

HP could also buy infrastructure vendors with an eye toward the cloud computing platforms of the future. With this in mind, Wang suggests integration and storage vendors as well as direct cloud computing service providers such as Rackspace.

Join the Stack Game

Despite all the hype about integrated and optimized stacks, I'm not convinced that the latest generation (borrowing from IBM's mainframe and AS/400 varieties) has gone much beyond the concept phase. From what I've seen, the term "integrated stack" has been greatly confused and conflated with the term "appliance."

Appliances are generally very targeted offerings aimed primarily at data warehousing. They are comparatively simple combinations of database, operating system, network connectivity and some level of analytic software.

IBM's Smart Analytic System is a bit broader, but as the name suggests, it's still confined to analytics, adding options for data-integration and BI software as well as a few analytic applications. Oracle Exadata is broader still, providing a starting point for data warehousing or transaction processing. But it's just a starting point. It's not like the middleware and applications are shipped in the same box, preconfigured, optimized and ready to go. Even Exalogic, the very latest from Oracle, does not include Fusion middleware, let alone applications.

Blunt Oracle's Marketing Spin

There are no independent tests or benchmarks that I know of -- or even market track records -- that prove that any single-vendor stack is any better-performing or more reliable than third-party combinations of hardware and software. Oracle now makes a lot of assertions about the superiority of its "Oracle Integrated Stack Testing," yet it has been less than a year since the company completed its acquisition of Sun.

HP, meanwhile, has had labs at SAP's headquarters in Waldorf for years where it does nothing but test the performance and compatibility of HP hardware and new SAP software. And there are already packaged combinations of SAP software and HP hardware aimed at rapid deployment for midsize customers.

HP and Microsoft are also working toward tighter integration, with the two companies having announced $250 million in investment toward co-innovation in January. HP has told Bob Evans that we should expect formidable competition on this front.

Is single-company ownership the only route to a tightly integrated stack? Absolutely not. With tight partnership, co-engineering and clever marketing (perhaps involving uptime and single-throat-to-choke guarantees), HP and partners could match (or at least blunt) rival claims about integrated and optimized stacks.

HP has been a partner of choice for both SAP and Microsoft. And even if HP invests heavily in SaaS vendors and cloud computing infrastructure, these alliances aren't likely to be in jeopardy. Coopetition is a fact of life in today's market, such that Microsoft is even partnering with IBM for hardware to run its Microsoft SQL Server Parallel Data Warehouse appliance.

Make Selective Acquisitions

Looking beyond enterprise apps and cloud computing, there are other acquisitions that would serve HP well. Buying Teradata, for instance, would bolster HP's high-end data warehousing business -- a market in which HP Neoview has failed to gain traction. As a bonus, SAP has been getting closer with Teradata over the last year and Microsoft has a longstanding partnership with the company whereby Microsoft BI serves as the front-end environment for Teradata data warehouses.

Informatica would be another good acquisition for HP, bolstering its data integration prowess and enterprise depth without really alienating the likes of SAP and Microsoft. EMC is another possibility. I'd defer to storage experts on the extent of overlaps on the hardware side; but EMC's VMWare, Documentum and other software assets are just what HP needs to bolster margins and drive services revenue.

In short, I'd advise Apotheker to invest wisely in long-term growth plays, to cultivate important partnerships and to not get too hung up on the tactical strategies of the day that worked in the enterprise software market of the past two decades. Times have changed, and HP can only look forward.Hewlett-Packard's choice of Leo Apotheker as the company's next CEO surprised many, but there are no surprises in the challenges that await him as leader of the world's largest technology company. HP must deepen its presence in software and services or it will face low profits, low growth and a marginalized position in the IT industry.