Separate Facts from the Virtualization Spin
 Joe Wilcox
News Analysis. Say that again, Microsoft? How many retailers are adopting virtualization? |
This morning, Microsoft dropped a press release that started off: "Close to three-quarters (71 percent) of U.S. retailers are turning to a broad set of virtualization solutions." Whoa, not so fast. That number can't possibly be true. The majority of U.S. retailers are small businesses. There's simply no way that most are adopting virtualization.
Microsoft has got to know this. Elsewhere the press release uses the qualifier number of retailers "surveyed." Later on the press release is more specific: "205 technology decision-makers at U.S. retail companies with annual revenues of $250 million or more." Those are pretty large-size retailers.
For example, according to the U.S. Census Bureau's most recent data (2005), there were 5.9 million business firms in the United States; 4.7 million have fewer than 10 employees. U.S. Census Bureau makes a distinction between firms and establishments, of the latter there are 7.5 million, of which there also are 4.7 million with fewer than 10 employees. This data doesn't include sole proprietorships and other very small businesses that generate revenue but don't record employee payroll. They are the largest segment of the U.S. business market, about four times the number of business firms. The point: Most U.S. businesses weren't included in the Microsoft survey.
I randomly pulled some data to show how small most of these retail businesses are. There are 6,236 retail bakeries in the United States, 4,292 are small businesses with fewer than 10 employees. The number of retail trade firms is 736,940, of which 585,497 have less than 10 employees. Cut another way, there are 1.1 million retail trade establishments, of which 680,987 have fewer than 20 employees. Most of these businesses don't come anywhere near $250 million in revenue. The majority don't even have IT organizations.
Microsoft sponsored the survey, which KRC research conducted. I'm not familiar with the analyst outfit, even though it's Washington, D.C.-based. Until six months ago I lived three miles up Connecticut Ave. from the city's border. My familiarity is really immaterial. The point: Sponsored surveys are just that sponsored. The company commissioning the research often has some agenda, particularly for information intended for public disclosure.
Microsoft wants to make a big case for virtualization, given what Hyper-V adoption could mean for Windows Server 2008 sales.
The press release is just another of those PR shell games that Microsoft and some other public companies like to play. Microsoft has talked about having greater transparency, which would be great. How about the company show more transparency—and honesty—in its press releases? Citizen journalism may be the hallmark of the new millennium's first decade. Microsoft's approach to big PR spin just doesn't fit with the rapidly changing way people consume information outside PR news channels. Bloggers' posts and IT managers and enthusiasts posting to forums, Facebook or even Twitter can be most unforgiving of spin.
Create, Communicate, Collaborate with IT Professionals at Ziff Davis Enterprise IT Link.
Comments (7)
"The press release is just another of those PR shell games that Microsoft ... like to play."
@Joe :
You JUST came to this conclusion joe?
A little late aren't you?
Posted by n0neXn0ne | April 29, 2008 12:48 PM
Any discussion about virtualization should include what it is used for by most computer users, and why it would also be useful to some windows users. You know, the basics.
Virtualization is most useful to Linux, BSD, Unix, and Mac OS X user. The number one OS to be run in a virtual machine is Windows XP. Windows XP programs and games that don't work in wine, crossover, or cedega, will generally run in Linux with XP in a virtual machine. This is great for some users who just cannot seem to use the linux equivalent program.
Now running Linux in a virtual machine in Vista or even XP makes little sense. Linux is vastly more stable than even XP, which was Microsoft's most stable release. Why use the "unstable" OS to run the stable one?
Now, if you can make a VM program that will install the client drivers (XP) so that they work on the host OS (Vista) which would not have a printer driver, but XP did, then this would be useful. However, Virtual Machines Programs actual work the other way, being the host OS has to have the driver, has been my experience.
Now I can also see the benefits of using Linux as the host OS and XP as the client OS on the internet. Playing online games that only run in XP for example. Programs/games that will run in XP using say the free program VirtualBox, will let you do that. Even if XP becomes infected, it will not infect the host OS, Linux. Also, if one just backs up the VDI file (XP) before the infection, then a simple delete of the infected XP VDI file and a copy and paste of the backup, and wham, you are back in business.
However, for Microsoft, I do not see the advantages of running an OS in a virtual machine with Windows as the host. In fact, what I do see is that Microsoft has bought out some good Virtual Machine companies in the past, to basically kill them off. MS is afraid of virtualization is my opinion. Virtualization and programs like Wine, are making the switch very easy to Linux. You can get virus problem free GNU/Linux at distrowatch.com You be glad you did, instead of paying the Borg more money for Vi$ta.
Posted by chips | April 29, 2008 1:52 PM
The distortion may be even greater than you have portrayed, Joe.
How many retailers are there with revenues greater than a quarter of a billion dollars? How did MicroSoft select those included in its surveys? How did it identify the technology decision-makers?
Since there does not appear to be U.S. census data about "retailers" as a classification, one has to be extremely skeptical about MicroSoft's list. It almost certainly used a "convenience" sample of those about whom it already had data (sector, size, "technology decision" contact persons) based upon prior close relationships. Do you think that those with whom MicroSoft has an active relationship are representative even of all large retailers, or are they more likely to be "early adopters" of "the latest and greatest"? I am pretty sure that MicroSoft does not have information--or care--about MY small business, for example.
Participation in surveys is usually very poor and generally heavily biased towards those who have strong feelings about the topic of the survey. In the case of virtualization, those who are committed to the strategy are more likely to be among those with strong feelings than those who are not.
Posted by Pinball | April 29, 2008 1:56 PM
One qualifier worth mentioning is that virtualization is a really big category that also includes terminal services (application virtualization) and OS virtualization technologies like VMware, Xen and the eventually arriving Hyper-V.
When you consider that a very popular use of terminal services has been POS terminals (low cost, rugged and low operating services needs), the retail adoption of virtualization is fairly high for places with 3 or more registers, especially multistore retailers.
Not to imply that the press release is not really news or that it connects back to a benefit for Microsoft (either the point of the release is either they have completely saturated retail or someone else has).
Posted by Tommy Aquinas | April 29, 2008 3:13 PM
Wish someone would apply similar critique to claims like firefox's downloads... where they are obviously counting the fact that a user downloads and installs upgrades... but they make it seem like each download is another distinct user
Posted by uhura | April 29, 2008 4:19 PM
@uhura:
I think you just did.
Posted by Pinball | April 29, 2008 6:12 PM
Later on the press release is more specific: "205 technology decision-makers at U.S. retail companies with annual revenues of $250 million or more." Those are pretty large-size retailers.
Even if these numbers are double what they should be. This is still impressive.
You jumped on a single, stupid point. This might win you fans among the Apple and LINUX koolaid drinkers, but at the cost of your own credibility.
Posted by Grep You Later | April 30, 2008 1:50 PM