Is it a paradigm shift for IT or just another overhyped technology?
With respect to Johnny Carson and Carnac the Magnificent, the answer is "yes and no." The question: Is SaaS (software as a service) an overhyped idea that is not much good for anything beyond application delivery?
First we need to define our terms. SaaS and on-demand are not interchangeable. On-demand is really about how you supply technology to a user. It is a technology solution. SaaS, on the other hand, is a task-oriented business solution delivered in a timely manner. (I have to thank Andy Mulholland, global CTO at Capgemini, for the definitions.)
One answer is no, SaaS is not overhyped, because SaaS is already far more than a way to reduce deployment costs, although we may not realize it. I spoke with Treb Ryan, CEO of OpSource, a company that helps ISVs transition their applications to an SaaS model. The real revolution, Ryan says, is SaaS as an enabling technology.
What's the single largest SaaS deployment currently? Not Salesforce.com but eBay. After all, what is eBay but an online application used by millions?
"[eBay] has revolutionized the garage-sale industry," Ryan says. That's a funny way of looking at it, perhaps, but what he really means is that eBay has spawned a whole new industry that replaces garage sales and hobbyist magazines, and has created hundreds of new online businesses, all using the eBay platform as their storefront.
Applications such as eBay, Google, or even Amazon.com are dynamic solutions that are constantly being redesigned and upgraded to help businesses sell more by getting more value from customers.
BlueRoads Software is another example. It uses the unique attributes of SaaS to help companies manage their channel sales, by bringing everyone --from local VARs to a company's own sales force to company reps to big distributors such as Ingram -- all onto a centralized site.
But I also answered yes to the question of whether SaaS is an overhyped model. I asked Capgemini's Mulholland what it can offer the enterprise and whether he saw SaaS hitting some kind of brick wall in terms of its capabilities. Mulholland believes it will.
All any company does is buy something, add value, and then sell it. To do that, the law says companies must follow standard procedures -- so standard, in fact, that auditors regularly check them.
"That is where SaaS comes in," Mulholland says. But he asks, "How do you buy better or sell better or differentiate the way you do business?"
It's the 80-20 rule. Eighty percent of a company's operations are standardized, and for those it can apply SaaS, but the other 20 percent requires a company to differentiate itself. SaaS, by its very model, doesn't really supply the solutions here.
Mulholland points to Siebel and its Project Nexus as an exemplar of how to address that other 20 percent. Nexus uses metadata, rather than source code, to describe what the application should look like. This allows Siebel to componentize its products into reusable objects, which in turn allows its customers to design industry-specific variations of standard processes using a drag-and-drop editor, instead of rewriting code.
It's too early to tell whether SaaS is a paradigm-shifting technology. My guess is, however, that it is, not because it can dramatically cut deployment and support costs, but because it will engender new types of solutions, such as eBay, that weren't possible before.