HP: My Day In The CEO Interview Hot Seat

HP Board: Hmmm. On to services.

Wittmann: EDS was a good addition to HP's existing services. For a company that's happy about being in businesses with 5% margins, it's easy to argue for going further down this road. Perhaps buy Sunguard and Rackspace and go further into managed services. Buy McKinsey or Deloitte and be a real player in business process consulting. These will all be higher-margin businesses than many of HP's existing businesses.

The board needs to decide what it wants to be here. If your goal is to climb the Fortune 500 list, then making more low-margin devices is the way to go. If you want to increase margins, services is a good way to go. It's also going to make you seem a lot more like a strategic partner than you are when you're simply competing with Dell to provide boxes. And since you don't own any application software, you can play an important role for companies that don't want an all-IBM or all-Oracle solution. There's a lot of those people out there.

HP Board: OK, so what about software. It's not a big business for us. Should it be?

Wittmann: Software is only 3% of your revenue, and it's reasonable to argue that playing Switzerland to all software vendors has served you well. I'm not sure if I fully buy into the Oracle appliance vision. It feels like the company just bought its first hammer, so suddenly everything's looking a lot like nails.

But you can see value in the concept and value in the architecture of its Exadata system, for example. It's built on Intel chips, the Linux OS, off-the-shelf networking, disks, and memory -- and yet it's architected to perform very well. You can build these systems too. And even if Larry won't let Oracle DBs run your hardware (which I doubt – but you never know), you still have the likes of Microsoft, Teradata and Netezza to work with. You've already done a lot of work to create turnkey systems for certain SAP products and such. I'd do a lot more of that and I'd improve my application performance management story and virtualization management story. I'd think about buying F5 as part of the strategy too.

The problem with the appliance strategy -- and every IT guy who's gone down this path know this, particularly those who've bought a lot of security appliances — is that you can easily make all the right appliance purchases in their own right and end up with an incoherent system that either doesn't accomplish your overall goals or is an integration nightmare. One appliance for your largest DB needs may make perfect sense; one for every single one of your applications and you're either going to be dealing with just one vendor or creating an integration mess. HP should be the company that rationalizes the big picture that calls for lots of software from lots of vendors. In that sense, you could out-do IBM at its own game.

HP Board: Well, thanks for your time. Anything else you want to add?

Wittmann: Umm, sure. Free Tibet and legalize 420!

HP Board: We'll get back to you on that.