I had a chat the other day with Dave Duffield, a software industry icon who founded PeopleSoft and, more recently, SaaS company Workday. Why, Dave, I asked, are you so confident in the profitability and growth potential of the SaaS model? After all, there's a lot more red than black on the industry's SaaS balance sheets. I share with you Duffield's thinking on the sustainability of SaaS.First off, Duffield says, more companies will move to SaaS once they realize it's less expensive than licensed software. OK, that's the expected answer from a SaaS vendor. But what he said next was more intriguing.

SaaS has the potential to be highly profitable because it's "sticky," says Duffield. In other words, it can be difficult for a customer to move off a software service and to another SaaS provider, and even harder to move from SaaS to on-premises licensed software. When a customer signs on for a software service, it's also signing on for operating systems, middleware, servers, network connections, databases, and the talent that goes along with all that.

Yes, it's signing on for an IT infrastructure, particularly if it's making a big bet on SaaS, such as using Workday for human resources, payroll, and financials. "Once people have implemented Workday, they're stuck, in a very positive way, with our solution and the services we provide, " he says.

That's interesting, because it runs counter to a generally accepted belief that SaaS suffers from a high "churn rate." Customers not happy with a SaaS vendor can cancel a subscription and move to another, or take the business process in-house on licensed software. This happens quite a bit, in fact, and some financial analysts expect it to happen even more in coming months as customers look to cut costs and downsize their SaaS subscriptions (having already downsized the employees who used them).

Duffield's contradictory view is based on a few assumptions. One is the idea that as more companies become reliant on SaaS, they'll have fewer investments in the IT systems and talent to be able to easily extract themselves from a SaaS relationship and take the job to their own systems. (I can see this happening particularly with young, growing companies. If they start heavily reliant on SaaS when they're small, they're more likely to stay that way as they grow.)

The other assumption is that as SaaS grows from single-point solutions into bigger things, such as ERP suites (Workday) and development platforms and ecosystems (Salesforce.com), customers are more likely to develop long and meaningful relationships with their SaaS vendors, and less likely to walk away if the relationship hits a rough patch. (And with no in-house servers and systems to run their ERP, what's the alternative?) Follow this line of logic, and you'll see that as the SaaS model grows, it starts to look more and more like IT outsourcing.

But what about all of those Wall Street analysts scratching their heads over SaaS? Duffield waves them off. "The financial community doesn't get it yet," Duffield says. "They can't figure out how to value Salesforce.com. But it's eventually a very highly profitable model in our space where the solutions are very sticky."

One person's sticky might be another's vendor lock-in. In fact, I've heard murmurings from a few readers about Salesforce.com making it difficult to break subscriptions without some financial pain. But SaaS is a business like any other, intended to make its purveyors some money. And if SaaS vendors can forge deep relationships with customers, those customers will be less likely to walk away when things get...sticky.

(Ever thought about the other SaaS, storage as a service? Check out our new report, Cloud Storage's Top Uses.)