Infoconomist: Business gurus say that big mergers rarely work. Why should this one be any different?
Michael Capellas: Empirical evidence doesn't support the notion that 'big mergers rarely work'. If you look at any large financial institution, consulting company or oil company in the US, they are the end result of a major merger.
Mergers work when you have economies of scale and complementary products. At that point, scale and the ability to really differentiate at a higher level of the value chain become important. Part of what people have refused to accept is the fact that the computing industry is maturing to where we are starting to use more and more standard components, standard building blocks.
There are actually only five large-scale global computing companies left – IBM, HP, Dell, Sun and EMC. If you look strategically, each of those companies will have to go through a significant transformation. IBM is going to go through a big change, and they've set some targets to do some major cost-reduction models. Obviously EMC and Sun are going through both product transitions and business model transitions. And Dell, which has done pretty well in component pricing, has still to find its future engines of growth.
The whole industry will have to go to a more effective cost model – so having the leverage to drive costs down is hugely important.
On top of that, very rarely do you see two companies that have an almost perfect complementary position. Compaq pre-merger had fault-tolerant computing, which is not a market that HP was in. Compaq didn't have a very strong position in commercial Unix, while HP was number two and now number one. Compaq had the number-one position in industry-standard servers, but HP had struggled there. Compaq was trying to move into system management software, while HP had a leading product with Open View. Compaq had great mid-to-high-end storage but OEMed the very high end, whereas HP had the very high end through its XP product lines. Both HP and Compaq were trying to extend their services 'play'. We [the combined company] went from number nine and number 10 in the industry to number three. We've got critical mass. And both companies were looking to do some things to improve the economics of their PC businesses.
IC: So what is the grand vision for HP?
MC: Customers will continue to want to buy more complete solutions, as opposed to putting the individual pieces together. There are two companies that can do this: IBM and HP. So how do we compete against IBM? The answer is we will continue to use what we call best-in-class components in the industry. We will partner with Intel for microprocessor development – while IBM develops Power [its proprietary chipset]. We will partner with Oracle and Microsoft in middleware and databases – while IBM tries to develop WebSphere and DB2. And we will differentiate on technology where there is true differentiation and use the best of the rest of the industry.
IC: As you point out, almost everything HP has is based on either open or Microsoft standards, with the notable exception of the former Tandem business. What plans do you have for that unit?
MC: There's still a huge market for mission-critical applications that need proprietary technology. The key is to only put proprietary technology where the customer really gets the value from it. For example, HP-UX is still the leading product for a mission-critical data centre and we will continue to invest heavily in that proprietary technology because it provides a function.
As for Himalaya [a Tandem high-end server], this is a goldmine. It continues to be the leading product in fault-tolerant computing. The merger of the two companies just makes it even stronger because it allows us to have an even bigger presence. We fully support Himalaya and are trying to grow it and we're working on the interoperability of it with our other product sets.
IC: Most economists think there is overcapacity in IT, with the possible exception of storage. How does that affect HP's prospects?
MC: There's no question that a tremendous amount of capacity was added during the Internet boom. This means that some capacity needs to leave the industry. There will be consolidation in component manufacturing. The market will be left with fewer, stronger players – of which we fully intend to be one, because we have an awful lot of cost levers to pull. I'd rather be ahead of the curve than behind it. I think, given our balance, I would take our position. I think you'll find our competitors reacting to us.
IC: If this consolidation occurs, HP may be left with fewer competitors. Does that mean you believe you will be able to increase margins?
MC: Computing margins will not rebound back to the levels in the Internet boom days. The effective companies understand that the margin model has changed, and manage it aggressively.
I think there are three components to driving better margins. The first one is a more complete high-end mix, complemented with more services. You can always charge higher prices, of course, but for me the second component of margins is reducing costs. There's no question that the synergies from the merger – which are quite substantial – allow us to make a much more effective cost base. The third piece is leverage with our partners. For example, the ability to have one standard microprocessor line across the entire server line gives us a huge cost advantage.
IC: You mentioned the importance of services for margins. Is HP planning any further acquisitions in that area?
MC: Nothing major. We've got enough to swallow right now. But because we are a very large company we will continue to make some spot acquisitions, either where we have some geographic capabilities that we need to build out or where we don't have critical mass in a certain country. You may see us making a couple of small services acquisitions or software acquisitions to fill out some gaps – but they're minor.
From a product point of view we feel really good with our server and storage line-up. In terms of our services capabilities, we're in pretty good shape in Europe and the US. We might add some capacity in certain segments of the outsourcing market where we don't have as much capacity as we may have market opportunity for. In some countries – particularly in Asia and South America – there are some markets we can fill in. But really that's about it.
IC: Would you consider something on the scale of HP's attempted merger with the consulting arm of PwC a couple of years ago?
MC: No. First, we've got a pretty major merger that we're managing right now. Second, we're pretty comfortable in our services capabilities. We want to be the best in the world at IT infrastructure and customer support. We don't necessarily believe we have to be in the systems integration business, nor is it necessarily in our DNA to do that. So no, we're not looking at anything major.
IC: When do you expect the IT industry to recover?
MC: If you look at the fundamentals, there is pent-up demand. When you talk to customers, people are continuing doing planning. Macroeconomics is driving the IT industry. Companies are concerned about their profitability, they are concerned about world economics, and so as a result they are cutting back. Demand for IT in itself has not gone away. But what's happening is more macroeconomic-driven.
I don't see any great second-half  recovery, and it's still unclear to me when the global economy turns. Will it? Of course. When it does, will IT spending rebound pretty robustly as a share of capital spending? Absolutely. But when that is, I really don't know. I wish I did.