Wednesday, January 24, 2007

SaaS as a Business Model

Over a few beers earlier this week, I got to talk about SaaS with a die hard believer. His argument went that the next generation of users have accepted and actively use software on the web. Thus, as they move into the workforce, they will not only accept SaaS, they will prefer it.

Not unreasonable. This has happened before -- Apple dominated the early education and home computer market because they put a few free machines in every school; AutoCAD became the standard for CAD/CAM because they made their software easily accessible to students and schools. Microsoft currently does this with cut rate pricing for education. The argument schools should use open source because it would cost $500 to outfit a PC with Windows+Office is silly as a result -- Microsoft sells media-less Windows licenses for $25 and Office for a few dollars more.

However, there is a catch to this. The reason these programs have succeeded is because they still met the basic requirement users have of their software: ease of use and scratching an itch. It is here that SaaS starts falling apart.

It's hard to get SaaS to beat a desktop equivalent, even with technologies like AJAX which push a lot of the hard end user experience work to the client so response times are quick. I'd argue that if SaaS for a particular kind of software succeeds, that says more about the poor quality of the desktop equivalent than it does about the SaaS.

"All mail clients suck. This one just sucks less."
-Michael Elkins, project creator of Mutt

In the case of, the poster child of SaaS, the CRM software that it replaced really was terrible. Customers have long complained about the end user experience of CRM packages. They are slow, were written by people who would never use the software themselves, and very costly to install and maintain. By comparison, offers a relatively straightforward interface, is incredibly cheap to start with and reasonably priced to grow with. If you need customization, they'll do it for a price. If what you want is useful to others and are willing to let them own the result, you can get it for cheaper. For a company that wants nothing to do with CRM aside from using it as a tool, this is an ideal situation.

But now let's look at online calendars, spreadsheets, and word processors... I'd link to a few of them, but besides Google, I expect most of them to have URLs that will disappear. (If that's not foreshadowing, I don't know what is.)

Online calendars are okay. I use the Yahoo one, but they are really behind the curve and I've just been too lazy to get authenticated SMTP on my mail server at home so I can switch to Outlook. (Or let it go and pay for any number of services that do mail hosting.) Others are better, but not significantly so. Their challenge is that they have to compete with Outlook's calendar which is actually quite good. It's easy to use, responsive, and integrates easily with email and tasks. For a group using Exchange, group functions are a breeze. Really, as an end user, I don't have pain with Outlook.

Online word processors and spreadsheets are still in their infancy and can't compare in either response time or usefulness. This can change, but it's going to be hard to compare with their desktop counterparts and the argument for price is weak -- if I'm that hung up on pricing I'll just get Open Office. Most importantly, even today's teenager has a notion of "their stuff" and having "their stuff" hidden away in some random computer on the Internet isn't nearly as satisfying as having the file on their laptop and being able to point to it. It will take a long time and significant strides in ubiquitous Internet connectivity before that changes. One only need to look at banking and the number of years it took before people were used to the idea of having money they never see. Even with the current state of banking, the gold coins business is still doing well because people like to see and hold their money.

Moderate successes are out there. Email as a service is doing well for personal use because most people consider their email as temporary to begin with. Those that have a need for heavy use or long term storage of their inbox use desktop packages and pay for the appropriate service. Email as a service for business users is still a very small market. Businesses that live and die by their email will pay for a service up to a few people, but it doesn't take many users before buying your own server becomes a better choice.

Coming back to SaaS as a business model -- I believe it can work. The caveat is that it will work for specific software niches where desktop equivalents aren't up to par. Until a SaaS implementation can meet both the end user experience requirement and scratch a real itch, it simply won't see wide spread success. There is still a lot of opportunity in SaaS and I periodically see a new application emerge that merits the business model. However, I doubt that it will achieve true widespread use anytime soon.


Post a Comment

<< Home