VP and GM,
HP's Business Intelligence Group
It hardly seems fair to interview a senior executive at a massive IT hardware, software and services provider barely a half year into the job. We'll make an exception in this case because Ben Barnes did not arrive at HP an unknown quantity. The new VP and general manager of HP's blossoming Business Intelligence Group is a seasoned veteran of startups and acquisitions as well as BI's big leagues. He knew Tom Hogan, HP's top software executive, from their days at IBM when Barnes managed what became a billion dollar solutions business. He'd known HP CIO Randy Mott during his early days at Teradata. Reunited with old friends and with the blessing of CEO Mark Hurd and Technology Solutions EVP Ann Livermore, Barnes rather suddenly holds the key role for Neoview, HP's just-launched, high-end data warehousing/BI platform. Combined with a new service arm based on the acquisition of Knightsbridge Solutions, it's a huge resource commitment and a steep ramp, as DM Review Editorial Director Jim Ericson recently discovered.
DMR: Neoview went from stealth to launch just a few months after you joined HP. How did you feel about a new Global 2000 platform in a fairly saturated high-end market?
Ben Barnes: The word "stealth" in this case means a tremendous investment in time, dollars and intellectual resources working with key customers throughout. If it was anyone else entering the market, I would say maybe the train had left the station, but as part of one of the largest technology companies on the planet, with a CEO, CIO and myself attuned to this space, I can guarantee that HP is not late to this market.
DMR: What's the Neoview proposition?
BB: We stepped back and asked what was wrong with the high-end machines of today. The first thing is cost. If you go out and ask high-end buyers, they'll tell you they're way too expensive to own and maintain. Second, they're not designed for the evolving enterprise data warehouse where we're talking about thousands of users, complex queries, concurrent data loads and querying at the same time. That implies reliability, availability, serviceability that looks more like a machine that runs the stock market than a machine that does queries.
DMR: Won't you need to establish a track record for those promises?
BB: The machine is based on the NonStop operating system kernel, but we've taken the database and tuned it for decision support as opposed to online transaction processing. The whole purpose of that is to bring to market a new machine based on industry-standard technology and hardware with a price point that is dramatically different that what we've seen. It is going to scale from the low end, a Netezza-type appliance to a Teradata-sized machine, except at a price point that looks a lot more like an appliance price point. I'd add that appliances are fine for specific workloads; they don't take on complex queries or hundreds and hundreds of terabytes or thousands of users loading and querying data at the same time. We are moving the puck in the area of more real-time applications built into business optimization usages. Some of this is vision, some is reality, but Neoview is certainly not a vision. It has already won the [government IT user conference] FOSE 2007 award for enterprise systems.
DMR: HP has always had a huge services arm, but not in data warehousing and BI. Was the Knightsbridge acquisition critical to launching Neoview?
BB: I certainly think so. When I was contacted about this opportunity, I can remember seeing a glimmer in Ann's [Livermore's] eye when I mentioned that services were going to be an important component of the project. I had a large service team in my unit at IBM and knew these were often consulting-led engagements. If you're going to try to build this organically, it would have been slow and painful, but HP was obviously well ahead of me. This is not a setting where you just deliver a machine and say goodbye; there are a number of service and application components. I don't think [HP] would have done Knightsbridge if they hadn't done Neoview and vice versa.
DMR: How big is the service component?
BB: Generically we've always had a big service business, about $16 billion for HP today. That's where Ann's star shined initially, when she was running the service organization. The services component for BI came from recognizing that we wanted to enter this market in a big way. We now have about 3,000 people globally at HP who do nothing but data warehousing and BI. If you rolled the clock back just a year ago, you would have been lucky to find 280 people.
DMR: It sounds a bit like IBM buying PwC Consulting.
BB: People do bring that up, but the reality is that this is the way most enterprise customers want to buy things. They like the idea of a service/software/hardware company working together, that infamous "one throat to choke." The difference is that we have a very strong partnering strategy by choice.
DMR: In a similar way, though, the partner focus has been very visible in your announcements.
BB: The theory is that most companies will have a proprietary stack that is locked in. Our strategy is to go after some pieces of that pie and keep good partnerships as we have with, among others, Informatica, SAS, Business Objects and even Hyperion, though they are now owned by Oracle. It extends to Oracle as well, even though the Neoview platform itself has its own decision support-optimized database that doesn't use Oracle. We want partnerships with lots of people in the marketplace, and that will help customer choice. But somebody always gets around to the question about all the consolidation in the market, so I'll ask the rhetorical question myself: "Whom are you going to buy next?" Well, I'm not here to announce anything.
DMR: So you're not contemplating a "soup to nuts" BI platform as is, say, Oracle?
BB: First of all, we have a great partnership with them. We've spent $6 billion on BI and some of that applies to Oracle, some of it [applies to] SQL Server, some to DB2. Our Neoview product line at its lowest level is going to intersect and overlap with some of our partners, and some people will make a point of that, but overlap is not the issue. The reality is that Oracle and DB2 are general-purpose databases that do wonderful things in specific decision-support, high-volume places with vast numbers of concurrent users, many terabytes of storage and very complex queries. You need a machine that can address that, which is precisely what Neoview was built to do. We also have about $16 billion in cash on our balance sheet. We'll keep our antennae up because a lot of smaller companies have interesting technologies that frankly might be very important.
DMR: Speaking of that, what have you learned so far about HP's intellectual assets?
BB: In addition to my general management job with Neoview, I've spent a good amount of time learning about a number of projects we are pursuing in the broader BI context. It has been very interesting and exciting to learn specifically about the work that has been under way at HP Labs. We've been able to identify several promising projects that are currently being pursued by about 100 people in California, Russia, China and India. These folks are working with very advanced techniques, and some of their ideas will be applied directly to the Neoview product and roadmap, and some will address other products such as SAP's BI Accelerator. Another example is a program we are working on with Accenture and SAS that is focused on mortgage processing. Some projects are moving into areas of advanced analytical mathematics that provide information in a decision-support environment and deliver insights from large data sets by looking at variables and constraints with linear programming and other techniques. It is pretty encouraging just to have the luxury other companies don't have - 100 people doing research into the next big thing.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access