Remember time sharing? Your company bought X number of hours on some distant mainframe (sometimes for heavy-duty processing in the middle of the night) and sent your data via phone lines. It was a big deal in the 1970s.
As soon as companies could buy software licenses and have programs customized to their specs to run on their machines, they did. Processing on your own schedule and in control of your own equipment became more and more popular.
Software as a service (Saas) is a different twist, in that you or your company gets to your data and programs via the Internet. You don't have to worry about updates (the software provider does that) and the software is ready and waiting when you need it. You pay for what you use. No more site licenses or worries about incompatibility with different computers. Oracle and others are busy helping get SaaS programs off the ground. Some media are hailing the environmentally-friendly aspects of SaaS.
But what happens when an outage prevents you from using the service, as DM News reports happened to Salesforce.com not long ago? The outage lasted less than an hour but Salesforce's users must have been biting their fingernails every second, unsure when access would be restored. And what happens when (not if) a SaaS provider raises its price? What options do users have if they want to quickly, safely, securely switch? These kinds of questions are likely to keep SaaS from widespread acceptance for now, IMHO.