VentureBlog has a post by David Hornik:
As fantastic as it is having greater predictability of revenue, I think that the product benefits of software as a service are what make the model truly compelling. Take, for example, PayCycle, a small business payroll service I invested in some time back. When I first invested in PayCycle, the team was just building the first version of the service. They had to make tough decisions about what states to roll out first, what features to build for each state, how to implement direct deposit, how to automate Federal and State tax payments, etc. With each new release of the software, the service reached more people, had richer features and was easier to use. Now tens of thousands of small businesses use PayCycle to do their payroll every month and they can bank on the service getting better and better as the development team continues to mature the service. The same story could be told of Salesforce, Facebook, Typepad, YouTube, NetSuite, the list goes on.
The beneficiaries of the service architecture are the end users — simply log in and get a better service than the day before. But this capacity to rapidly innovate also leads to greater revenue as attrition is driven down to single digits and lower over time. Ultimately, software as a service is a win-win for end users and startups alike. I have no doubt that I will be investing in interesting software services for years to come and look forward to watching the ones in which I have already invested continue to thrive.