Research firm Gartner forecasts that by 2011, 25% of new business apps will be delivered via SaaS, a jump from 5% in 2005. So while software-as-a-service is in its infancy and the biggest bucks still go for traditional on-premise software this new delivery paradigm is growing up fast.
As the software industry moves toward SaaS, one of its defining questions is: what is Microsofts approach to this new delivery model? As one of the 800-pound gorillas of the software business, Microsofts decisions tend to have industry-wide repercussions.
To find that out, Datamation spoke with Tim OBrien, director of Microsofts Platform Strategy Group. In a wide-ranging discussion, he spoke about the companys early SaaS initiatives, and its emphasis on what he refers to as a software plus services approach.
OBrien notes that many industry observers view the movement toward SaaS as an either/or proposition: Either businesses will host their software on-premises, or theyll move overwhelmingly to a SaaS model.
But he disagrees with this dichotomy. Its not either/or, its about and, he says. This and scenario is where the industry is headed, in his view. Businesses will combine SaaS applications with on-premise applications, choosing between them based on a wide array of variables.
Therefore, he says, Microsofts strategy doesnt focus exclusively on either the on-premise or the SaaS model. Instead, the emphasis is on facilitating customer choice.
If the need is keeping that application on-premise and customizing it, and extending it, and doing all the things that you do when you host an application in-house, well help them do that, he says. However, If they just want to move costs or overhead or some commodity outside their business and put it onto the cloud that is, host it where someone else can worry about it we want to be able to do that, too.
OBrien urges IT decision makers to avoid viewing the difference between SaaS and on-premise models as a black or white issue. In reality, he says, its a question of trade-offs.
What is the goal youre trying to achieve? Is the goal taking a commodity business process and moving it out of your environment so you can focus on things that are more core to your business and value-added? This goal could call for SaaS.
Yet the total costs of SaaS need to be calculated, he says. Think of renting software. Thats what software-as-a-service is in terms is pricing models youre subscribing to that. So you have the whole lease vs. buy discussion with the car dealer. If youre going to drive this thing until the doors fall off, you should buy it. If you get a new car every three years, you should lease it.
Indeed, the car lease vs. buy analogy calls to mind a key issue with SaaS. The canonical question is: at what point have you paid enough for your software? So if its an economic discussion you need to look at, over time, what is the cash outflow stream for service-based procurement going to look like, versus if I buy the perpetual license and pay maintenance?
To be sure, OBrien sees the advantages of the on-premise model. Theres a whole bunch of capability in terms of processing power and storage that lives on the edge of the network, he says. And to just say, Software as a service is going to be the be-all, and kind of the endgame here, ignores all that unused storage and processing capacity that lives on the edge.
Furthermore, what about those moments rare, but not gone when people arent connected to the Net?
Offline capability matters, he says. Despite what you read about the ubiquity of broadband and wireless access, there are some situations where you just dont have access to a network.
There are situations where you need local access. There are situations where you want to take advantage of some of the power you have on your local desktop. And there are some situations where you dont want to put sensitive data up in somebodys cloud. And there are regulatory issues there as well, like HIPAA, which says you cant store patient data in a third party data center.