Microsoft's recent Online Services announcement was just the latest in a long measure of hype around the term "Software as a Service," or SaaS, in the information technology market. SaaS describes a delivery model in which customers get access to applications on an on-demand basis, without having those applications deployed on their networks, and without paying any up-front capital expenses. The key differentiator between SaaS and other versions of hosted software is that a SaaS application is designed to run on the Web, in a hosted environment.That unique design and delivery can have its advantages, but I find that people tend to use the term "SaaS" to simply mean "hosted software." To hear some people talk, you'd think SaaS is going to change the world. But while software that's optimized for Web delivery makes hosted services more cost effective and, therefore, appealing, hosted services themselves hardly represent a new idea-and there's no reason to think that the delivery model will be any more successful now than it has ever been (down economy and all).
One reason early hosted models failed is that they couldn't scale-providers were housing rich applications in their datacenters, in a separate implementation for each customer, making it difficult (indeed, for most providers, impossible) to make the numbers work. With thin, web-based clients that are delivered in a multi-tenant architecture, the hosted delivery model can be profitable, as the economies of scale are finally realized by SaaS providers (and, presumably, passed along to customers).
There are other benefits to Software as a Service, although none are different from those of other "hosted" offerings. These include fast and easy deployment; the elimination of ongoing maintenance fees and lowering of operational costs; the reduction of bugs and patch fixing on the applications themselves; and the ability to scale up or down to the right number of end users almost immediately. It's also easy to switch providers and applications, at least from an IT point of view (business users may not want the "flexibility" of using new and better mousetraps every month).
From the vendor's perspective, SaaS offers benefits as well: SaaS providers maintain control of their software and can protect against illegal use; they can update versions as needed, quickly and in response to customer requests; they can limit the number of bugs inherent in the software; they can ensure all end users are always on the latest applications; and they gain an ongoing, reliable revenue stream.
But there are downsides to SaaS as well. One is that the applications cannot be customized for a particular organization--to meet unique business needs, or fit specific business processes. So there is no competitive advantage available in the software itself-and that makes SaaS a delivery method most suited to applications that don't shape or depend on a company's unique business or processes (think HR applications, conferencing and communications, and office productivity).
Also, it is virtually impossible to integrate SaaS and on-premise applications, making SaaS a bad option for companies that need to support and connect multiple applications (say, a conferencing system and an ERP system), or that need to access backend corporate data within the SaaS environment-at least not without strategic relationships among the various vendors. It's also harder to embed communications in business processes (CEBP) when the communications are delivered as a service. These challenges can be mitigated somewhat if the SaaS vendor uses common APIs, and most SaaS applications can handle data exchange in batch processes.
Finally, while lower costs are often touted as a reason to use SaaS, it is often the case that companies pay more for hosted applications than they would to simply buy the software once, and then maintain it over its lifetime. Even with the most competitive pricing, after about a year's use most companies will spend more per user on a hosted basis than if they'd bought a user license. Many companies are willing to accept the higher costs over the long term in exchange for eliminating an initial capital expense, but software leasing arrangements can also lower up-front costs. Other companies absorb increased costs because SaaS saves them from having to hire people with the skills and expertise to manage the applications on an on-going basis-and to be sure, that can save significant amounts in ongoing operational expenses.
None of these benefits or costs are new, however. And this isn't the first time hosted services have been available during a down economy. So I'm not sold on why 2009 is going to be the year IT managers decide to ditch major investments in on-premises applications in favor of hosted services.
To complicate matters further, we now have "Cloud Computing" as the next big thing in software delivery models. What is cloud computing? The Wikipedia definition reads: "Internet ('cloud') based development and use of computer technology ('computing'). It is a style of computing in which dynamically scalable and often virtualised resources are provided as a service over the Internet." Sound familiar?
For more on this, customers can download Frost & Sullivan's new market insight at www.frost.com.
