VMware is approaching the kind of turning point that defines a company as either a brilliant success or a flash in the pan, and it’s not entirely clear which way it’s going to turn.
VMware has been king of the virtualization world for a long time and deservedly so. It didn’t invent virtualization, but it certainly made it practical in distributed environments and did more than anyone else to popularize it.
It also built up the hunger for it among user companies to the point that a July re-release of a Yankee Group report (originally released in February, and updated at the end of last month to keep up with dramatic changes in the market), estimates that 75 percent of all U.S. companies are either using virtual servers now or plan to do so in the near future.
That fits with other reports—both quantitative and anecdotal—but the more telling bit is that 40 percent of users surveyed plan to use virtualization products from more than one vendor.
Again, in a macro way, that’s not all that surprising. The same thing happened with networking hardware, networking software, operating systems, high-end server hardware, network and systems management software, VoIP systems, email, databases, Web serving and nearly every other area of IT. First there’s a pioneer and everyone uses its stuff; then there are imitators and users tend to pair off with one vendor or another; then users get tired of all that nonsense and demand the vendors support one another’s products so customers don’t get locked into the limited options available from one vendor, then all the weaker players get bought up or drop out of the market, leaving the strongest companies to compete using their ability to seamlessly integrate products from many companies, manage those products, and build on top of that a few top-end functions other vendors can’t match easily.
It’s a consistent pattern, but it’s not a comfortable one for either the pioneer or its most dedicated or early adopting companies.
Right now, in Yankee Group’s estimation, VMware is at the top of Little Big Horn wondering where all those Indians came from. That’s not completely fair to VMware, considering that Custer was attacking against direct orders as well as common sense.
VMware isn’t attacking anyone directly, but it’s not doing much to defend itself or pick a new battlefield, either. It’s just sitting there like a Chinese emperor relying for the Great Wall to defend the empire against Mongols who bribed their way through or disassembled their way over remote portions of the wall on the way to Peking.
China only won by accepting the Mongol conquest and civilizing the conquerors—changing the nature of the contest (probably unintentionally, it’s true) by absorbing the invaders into their own culture.
In the computer business, Embrace and Extend (Smother and Subsume) is better assumed as a tactic for Microsoft to use against others, not one for others to use against it.
No matter how sincere its effort or pure its heart, it’s hard for a manatee to smother a whale.
But VMware has a lot of advantages, but it’s far from having a lock on everything having to do with virtualization, according to Yankee Group analyst Laura DiDio.
“VMware is only the market leader in server virtualization. We are still in the very early stages of other segments of the virtualization market: desktop (VDI), application, storage and network,” she says.
Citrix is strong in desktop virtualization; Microsoft is stronger than you’d think in application virtualization, following its 2006 acquisition of Softricity, she says. VMware responded by buying ThinStall, which still trails Softricity in market share.
“It’s too early to call the network or storage segments, although VMware has to be considered very strong because it’s owned by EMC,” Didio says. “Overall the competition among the vendors is intense and cutthroat which bodes well for enterprises. It means that ALL of the vendors have to raise the level of their respective games and deliver great products the first time out, give their customers better deals (than they would have in the absence of the various rivalries) and of course, deliver great value-adds in the form of management, security and after market technical service and support.”
That’s a much more parlous situation than a vendor just waiting to see how successful a rival will be with one product. It’s a demand for either a major change of course or impressive acceleration.
The key will be to give customers who will be using Hyper-V or XenServer or other competing products anyway to do it as easily as possible. To succeed, it must be clear that VMware products have been designed to manage all aspects of a virtual infrastructure, especially if parts of that infrastructure are short-bus versions of VMware’s own products from other vendors.
VMware needs not only to support products from Microsoft, Novell, Citrix, Virtual Iron and others, but do it so overtly and so well that Microsoft customers feel the need to explain why they’re using Microsoft management products (which Microsoft promises will support VMware) even for Hyper-V. If they have to explain why, no matter what the merits, they use the more-expensive VMware products when there are cheaper alternatives available, VMware loses ground.
I’ve heard all the arguments about why VMware’s greater capabilities are, in fact, cheaper than having to kludge up a way to do similar things with feature-deprived products from other companies, by the way. There are still people, whether ill informed or not, who will put virtualization managers in a position to defend their own product decisions, largely based on price.
These people will often sit in the vicinity of the CFO’s office, which makes them both hard to explain things to, and impossible to ignore.
Microsoft inevitably wins this argument unless the benefit of using VMware to manage other company’s products is so clear and so easily cost justified that an IT manager can explain it to the CFO with one paragraph’s worth of single-syllable words and absolutely no three-letter acronyms.
Managing multivendor virtual infrastructures would be a major benefit, but VMware could also take advantage of all the storage and consulting and integration expertise of parent-company EMC to build itself out as a full-service data-center supplier whose particular strengths are the efficient use and management of IT resources through server and storage virtualization and management.
That’s a mouthful, but it’s a strategy, not a product comparison and not a price comparison.
No matter how often it’s said its expertise is in data-center technologies or how many of its products actually work effectively within a data center, Microsoft has never been able to build a lasting case for itself as a replacement for IBM or EDS or any other build-me-a-datacenter services company, and probably never will.
VMware has a good chance to put itself in that position, building a firm future for its products and its customers in the process.
So far there are no clear signs that VMware (or, more importantly, Joe Tucci and the rest of the EMC board) have decided to do that. But I wouldnt be surprised if they had something like that up their sleeves. I’d be disappointed if they didn’t, in fact.
I wrote a few months ago that I thought VMware was trying to turn itself into Banyan, and I think it still has a good chance to do so. (I loved covering Banyan, by the way, not because of the arrogant aloofness of its management, but the uniquely bitter enthusiasm with which its customers supported it, even knowing it and its “superior” technology was doomed by onslaught of Novell’s relentlessly effective marketing machine.)
It would be a shame to see that happen, though, considering the track record VMware has already built up, the scope of the redesign it engendered in what is considered mainstream IT and, most of all, the emotional, intellectual and financial investment its customers have made in it.
VMworld, the company’s annual marketing-and-training extravaganza is at the Venetian in Las Vegas in September. Seems like a good venue from which to place a big bet on a major new direction.