Congratulations Moshe!

On my first day back from a New Year’s break, I got a surprising number of emails and calls about IBM’s acquisition of XIV announced today.  Thanks to a friend who saw a Globes report in Israel, I got a heads up on Sunday the 30th, regarding the pending acquisition, which gave me a couple of days to consider the implications before the official announcement. For the record, I have no idea how much IBM actually paid for XIV, and if I did, which I just told you I don’t, I couldn’t tell you.  Globes estimates $300-350 million, which, if true, isn’t a bad return on investment for the founders.  That much, I do know.

Bad news being more popular than good news, the first two questions are always the same:

  1. Who does this hurt?
  2. Which product is XIV meant to replace?”  

The two hypotheses I’ve heard are:

  1. IBM can’t make enough money selling the DS4000 and is looking to ditch OEM-partner, LSI.
  2. IBM’s DS6000 and/or DS8000 Turbo is underperforming and IBM needs a replacement now.

If I take a deep breath and consider the hypotheses, I’m left with a quick conclusion that short-term, both hypotheses are wrong, that no one and no product is hurt, except for maybe some third party that has a technology directly competitive with XIV, maybe someone who was also hoping to sell to IBM.  If there is such a beast, they will have to find another suitor.  I can name at least three other suitors, if you’re interested. 

I know it’s not very exciting to say “no one gets hurt,” but regardless of where XIV is intended to fit in the portfolio, I’ll bet it still takes IBM a good six months to prepare the field to sell, ship and support a product that is already IBM production-ready.  I’m also willing to bet that XIV, while probably very good, is not quite IBM production-ready.  Few things are.  “Ready for prime time” and “IBM-ready” are not the same thing. IBM will, I suspect, continue to sell DS4000/DS6000/DS8000Turbos, continue to invest in them, and within the normal product cycles that most IT products have today, the products will live a typical life and die a typical, relatively-graceful death.  My only caveat might be the DS6000, about which I have no objections, other than the fact that it violates the McArthur Rule of Scaling,  which states that “Gravity favors products that scale up over products that scale down.”  Several important and knowledgeable executives disagree with me on that “rule,” by the way.

In my opinion, the smart move for IBM with XIV, is to leverage the product as “additive” to the current portfolio, which it is, to help IBM address markets that they may not currently serve well, which it probably does.  Over time, they may add features to the product to make it more cost-effective to supply and more suitable for a variety of environments.  But that will take years, not months.

The opportunity that IBM has with XIV reminds me a bit of EMC’s early days, when the team around Moshe Yanai, XIV’s founder and the original Symmetrix architect, positioned the Symmetrix to solve a narrow set of IT issues, like the high cost of IBM mainframe memory.  Over time, Symmetrix became the solution for a number of environments and use cases.  A very big number. But over time.  Transitions take time. 

I still remember writing the letter that explained to EMC why my company couldn’t use EMC’s then-current version of the Symmetrix (the 4800), but when they addressed certain single points of failure, they should come back for a second evaluation.  I still remember the IBM boxer shorts my then-CEO received from EMC’s founder, with a note that said something to the effect that “These should feel comfortable, since you’re obviously in IBM’s shorts.”  I’ve got those shorts stored away, but will one day donate them to a marketing-missteps museum, if one ever exists.  The misstep set EMC back a year at the account.  I’m sure IBM won’t make similar mistakes. And Moshe’, I hope you like the fit.