I spent time with Gary Orenstein of Maxiscale today, and we spoke about what was going on with the Data Domain/EMC/NetApp wars – both agreeing it sure has been interesting.
Gary wrote this blog back in May, where he raises a very interesting point;
“Back in November of 2003, NetApp announced the acquisition of Spinnaker Networks for $300 million, a move aimed at getting a jump start in the clustered storage market. Five weeks later, EMC announced the acquisition of VMware.”
I forgot just how close in time those deals were. EMC has since done a billion acquisitions – and the ones that have had the most material impact, such as VMware and RSA, have been businesses that have been left alone. The VMware buy for $600M worked out pretty good for them, no? While they have acquired lots of “technology” that has been incorporated into the traditional EMC product lines, they clearly haven’t been reticent to acquire a “business” and try to grow it. NetApp has only had a handful of acquisitions. Spinnaker, as most of their acquisitions, have been around technology – and more specifically, how those technologies can be incorporated into WAFL (their O.S.) to add value.
The difference has been that with the possible exception of Decru, most of the things NetApp has purchased have been interesting features/functions that it saw necessary to advance WAFL. EMC saw discrete “businesses” that could leverage the brand, money, sales, customer base, etc.
If NetApp is to win the battle for Data Domain ultimately, it represents a clear departure from the “it must be WAFL” behavior historically exhibited. They can’t wait for WAFL to integrate DD. Instead, they must be viewing the synergies they could gain by simply applying their value to the DD product set. DD is an appliance with a specific function. It does that function really well. NetApp has proven it can take an appliance with a specific function and sell a ton of them. If it can do that with this NEW (read: non-WAFL) appliance then it puts itself in a position to do the other thing it has done really well – to sell incremental software value upgrades on top of the appliance platform. That’s how it maintained 63% margins forever, to the envy of all.
EMC is clearly better at taking a business and doing what it does – using the strengths and not applying the weaknesses. However, in this case, it remains to be seen if EMC would leave DD alone the way it did VMware and RSA (meaning they might put new teams in place and leverage the brands, but leave it as its own entity). Either way, there is a lot of product rationalization that has to occur – something that NetApp doesn’t really have to deal with. I don’t necessarily see how it could be kept at arm’s length since EMC has so much core value already in the space with Avamar, Quantum, and Falconstor based offerings - aimed directly at the core existing EMC client base. EMC does have Dell, which is a big wildcard in all of this, who can consume a lot of this caliber product, regardless of the packaging.
It seems like EMC might be forced into a “technology” buy whether they like it or not - and NetApp can finally take the opposite position. Interesting for sure.
For NetApp to pull the integration off successfully, it has to get out of its own “NIH” issues and go, go, go. It has to embrace the fact that this is NEW – and not WAFL based and not worry about integrating it (at least not on day one). I can’t help but wonder that, if this approach was taken with Spinnaker, if we would have seen products long ago. Here we are six years later and we’re still waiting for the Spinnaker/WAFL integration to play out in any meaningful way. The issue there was $300M. The Data Domain stakes are much, much higher. Dave Hitz told me I would see the Spinnaker stuff really kick in this summer and I hope he’s right, but I still wonder what impact a non-WAFL based product line might have had by now.
Final thought – where does Quantum end up in all of this? They have strong patents in the dedupe world. They have a banker named EMC with $100M or so in place. They are the only legit, large tape player outside of IBM – presuming that Oracle/Sun/STK keeps heading south in that market, they have global channels and brand, etc. I can’t see them going away for all of those reasons, and am starting to think that they are a short-money buy with some strategic possibilities. If growth can’t be organic, buying a huge revenue stream and channel – even if it’s only on TAPE – isn’t a bad idea. In the words of Data Domain, tape sucks – but it ain’t really gonna die anytime soon. Someone will have to serve that market as there is no way folks can let IBM just have it freely. If Quantum can use the technology and patents they own to do more damage in the dedupe space, all the better. I’m no lawyer, but since DD and EMC already license the Q patents, does NetApp screw with them by buying Quantum if it isn’t the successful bidder? HP maybe? This saga may have a long run!
We hope to have episode 1 of ESG-TV’s “Spotlight Series” up tommorow – it’s with Frank Slootman, CEO of Data Domain. We filmed it a few months back, but he’s great and, at times, even prophetic. He spends most of conversation talking about interesting parallels of successful newcomers battling big incumbents – and makes great comparisons to NetApp, Cisco, EMC, and VMware…
Share
