Showing that oversimplification isn’t unique to network reporting, there’s recent discussion that SaaS will have a major impact on the software market by making it harder to justify incremental functionality in “resident” versions of software to promote upgrades. In fact, Novell was the original poster child for the notion of “function value slip”; its network operating system could not create valuable new features beyond file/printer sharing and thus failed to generate upgrades, eventually pushing Novell to a second-tier player.
Productivity software is already showing a similar problem; fewer companies upgrade to each new version of Office, and it’s not that they’re adopting a competing product; they’re simply continuing to use what they have. The impact of SaaS is really the same as the impact of open source or freeware products: A lower-function version of a tool is fine if the area where functions are being lowered isn’t one a given user depends on.
One truth about the impact of SaaS is that the TCO for desktop software is increasingly the cost of supporting it, not the cost of the software, and that cheaper hosted models may be the way of the future. Whether enterprises would elect to host their software on somebody like Google, however, is another matter. Right now, enterprises believe they would be three times as likely to use their own private cloud for centralized application hosting. Google’s Chrome OS and Microsoft’s Gazelle project may be aimed at supporting these users.