HP Acquires EDS: More Cloud Computing Fallout?
Posted by Bob Warfield on May 13, 2008
Much hullaballoo in the blogosphere over this acquisition as one would expect from a major transaction like this. The pundits can’t decide whether it is too small to place HP on a plane with IBM for services (“They should have bought Accenture or one of the offshore players”) or whether it is too big and eye-off-ball for what should be HP’s main mission–software. From HP’s perspective (make that Hurd’s perspective), the deal is a slam dunk on the numbers. What that means is that it is obvious to the HP planners how to cut costs so that EDS becomes a profitable new jewel in HP’s crown.
As I sorted through all the write ups, I couldn’t help but be struck by the similarities with Oracle’s acquisition game and how this all relates to the Big Switch (Nick Carr’s term for a movement to Cloud Computing) for IT. I come away with a sense that EDS was struggling a bit, swimming against the current as it were. This is not unlike the legion of On-premise software companies that Oracle has snapped up. They weren’t exactly failing, but they were tired swimmers and it looks like the current will only grow stronger. EDS is very much the services equivalent of a Peoplesoft or Siebel in this sense.
When faced with a major paradigm shift, big companies have to decide whether to double down and invest in keeping up with the shift or whether it makes more sense to fold. Folding is advantageous for the biggest players who can scale through M&A and do so profitably by cutting from their acquirees. Cuts are straightforward once the mission changes from profitability to growth. Just think how many projects your own company has going on that are future-focused and haven’t reached critical mass. Nearly all of these will be ended together with the usual infrastructure cuts (we don’t need 2 finance departments, 2 HR departments, 2 CIO’s, yada, yada). So HP and Hurd, and Oracle/Katz/Philips can do what they do best and act as shrewd number-crunching consolidators. It’s all part of making the market relentlessly more efficient.
It’s not a bad deal for HP. Protestations about acquiring Accenture to the contrary, they now become the #2 IT Services provider. Apparently there is some synergy in the area of automating data centers–EDS was the largest customer for HP’s OpsWare suite. Automating data centers is a good cost reduction move for IT, but it still can’t compare to the 16:1 cost advantage SaaS brings. Companies will likely always have some activity requiring their own datacenter, but smart companies should be working to reduce that with SaaS as much as possible, hence there could be competition for a dwindling pie here at some point. And, as Jeff Nolan writes, there is also the opportunity to bundle in other HP offerings, especially hardware, with EDS’s outsourcing business.
Who knows, maybe HP can push this into a higher level strategy to get into the Cloud Computing game on their own. Certainly EDS has been very active in the application outsourcing business, but historically that’s been an ASP’s game and has not been very competitive against true SaaS offerings. Dana Gardner seems to feel this is exactly the right move when he says:
HP with EDS has now clearly staked its future on the top prize in IT: next-generation IT operations efficiency, proper outsourcing methods, cloud computing services management, and high-level consulting as the onramp. This amounts to business transformation via IT transformation via IT multi-sourcing.
I’m a little less enthusiastic. Being an admitted card-carrying member of the SaaS religion, this all feels like a bit of rearranging the deck chairs more so than actually addressing the Cloud Computing threat. I am respectful enough of Mr Hurd’s (and his lieutenant’s) expertise not to see this move as a bad thing at all, it just feels more short term/tactical/number massaging than big bet/long term/game changing.
Some folks I was chatting with saw this as another signal that the IT world has fundamentally changed, that IT is now a commodity, and that the future is what Oracle and HP are doing. In short, there’s little hope for smaller companies and innovation. Someone else in the conversation observed that most of the Big Enterprise plays made their bones with big customers. The SMB market cannot afford to do things that way, and it’s a huge market. SaaS can tap into it very effectively.
Whether the new generation is going after a fundamentally different market, or whether even the old market is headed for the new generation, I don’t see things as “total commoditization and less opportunity” at all.
One generation is mature, and a new generation is off and running, growing like weeds. I would hate to be at an on-premise perpetual license company right now. Even mighty Oracle eventually was unable to innovate, they were unable to take much share with their apps business, and so they needed another alternative. SaaS was closed—the business model was absolutely corrosive to Oracle’s existing model. What to do? Do what the last guys did when client server was eating the prior generation’s lunch. Follow the CA model. Become the hostel for over the hill software (and services it seems) companies. Get some smart investment banking or financial executives on the team and you’re off to the races.
The real question here is what happens when the M&A feedstock runs out? The consolidators are in a race to keep their M&A currency high (PE or Sales/MktCap, your pick) while acquiring stalled companies with lower valued currency. Will it enable them to transition to SaaS by buying their way in through acqusitions? Way too early to tell, but these guys are stuck in the cross hairs between open source and SaaS any way you look at it. The good news is they’re really smart and have played the game incredibly well so far.
It’s funny to ask people who the big software companies were before Oracle, SAP, et al. Huge empires can vanish in the mists of time surprisingly quickly when the new model is compelling enough. Why wouldn’t SaaS and the rest of the Cloud Computing montage be at least as compelling as the Client Server wave that made Oracle what they are? What we’re principally missing so far is the double witching hour of growth acceleration that included Y2K and the Internet Bubble. Perhaps not having those urgent events will provide enough time for the old school to morph gradually enough this time.