Gartner analyst Lydia Leong was kind enough to respond to an email I sent her regarding this debate over Gartner’s $46 billion cloud computing market.  To be clear, I enjoy Ms. Leong’s CloudPundit blog and generally believe that she gets cloud computing.  She was kind enough to respond to my email with her thoughts up on CloudPundit.

Here is an excerpt of my reply:

Thanks for responding.  I have access to the press release - not the full report.  The press release disputes what you say.  It talks only about IT services and not consumers (the words “consumer” or “consumerization” are not used in the release).

While I get the point about some traditional IT services being advertiser supported, this represents a very small fraction of both advertising and IT functions.  Corporate email moving to the cloud (or more typically, SMB and nonprofit email), can be ad-supported and this qualifies in your offset.  However, most of Google’s, Yahoo’s and Microsoft’s ad revenue comes through either their home page search or through Adsense or other types of placement on millions of web sites - most of which provide no IT-related service.  Also, auction-based pricing for advertising (e.g. AdWords) is not nearly all online advertising.  Much is CPM-based and these types of ads are also used to support Web sites with IT services being supported.  In fact, selling ads on a CPM basis is typically far more profitable for the service provider than AdWords.

If you’ve seen the ad business lately, you’d know that it is virtually impossible for most businesses to live on ad revenues and provide a reliable and valuable service.  Most cloud services will be paid for through any of a number of pricing models, but advertising won’t be a significant percentage.

Further, the press release makes a SaaS/cloud distinction around “true multi-tenancy,” which is not how advertising auction systems work.  These are traditional Web apps that do not need to adhere to multi-tenancy models because the users are not able to modify how they operate or extend them intrinsically.

By lumping advertising, e-commerce, payments and online HR services into your definition of cloud computing, Gartner dilutes the term into a state of total irrelevancy.  What you really seem to mean when you say this is cloud = every internet-delivered service that remotely touches corporate IT.

Gartner’s definition of cloud computing is at odds with the U.S. Government (http://csrc.nist.gov/organizations/fissea/2009-conference/presentations/fissea09-pmell-day3_cloud-computing.pdf), McKinsey (http://uptimeinstitute.org/content/view/353/319), IDC (http://blogs.idc.com/ie/?p=190), U.C. Berkeley
(http://www.eecs.berkeley.edu/Pubs/TechRpts/2009/EECS-2009-28.html) and nearly every other definition that’s out there.  It provides no actionable value, which in the end is where Gartner tends to shine, and it’s not defensible in any discussion with practitioners and other experts.

I had asked a few people at the Gartner outsourcing meeting if they agreed with your definition of Cloud Computing.  Most said no, some snorted derisively, and one commented that “Gartner’s late to the party as usual and isn’t dressed properly.  They just don’t get the cloud yet.”  These were all Gartner clients, not vendors.

“Is it confusing to see the giant number with advertising included? It can be. I often start off descriptions of our forecast with, “This is a huge number, but you should note that a substantial percentage of these revenues are derived from online advertising.” and then drill down into a forecast for a particular segment or subsegment of audience interest.

Giant numbers can be splashily exciting on conference presentations, but pretty much anyone doing anything practical with the forecast (like trying to figure out their market opportunity) looks at a segment or even a subsegment.”