I have been seeing a lot of blog posts recently about the current state and future of SaaS. Matt Tucker makes an intriguing argument about why single tenant is needed (and feasible) for SaaS based on the public cloud. Bob Warfield (Update: my apologies Bob, I must have had the Miami Dolphins on the brain) weighs in on the economics (economies of scale) of SaaS that argue for multi-tenancy instead of single tenancy on the public cloud. Finally, Josh Greenbaum believes that the debate between single tenancy and multi-tenancy will ultimately be moot (though I have to say I am not in love with the SaaS 2.0 term that Josh uses -- anytime someone coins a "2.0" anything I have an immediate negative reaction). Josh believes the debate is going to shift from SaaS being better, faster, cheaper to SaaS allowing the creation of new applications whose value proposition is defined by their ability to aggregate data and processes (and stakeholders) in the cloud.
In a sense, all three perspectives are right. It really depends on what business problem you are trying to solve and, as a vendor, what your value proposition is for customers. Some customers want solutions where they have the most control and can customize it to meet their specific needs. This is especially true if it is something that is strategic and competitive differentiating for the customer (and they would prefer to not build it from scratch). Different customers may choose to take that core solution and go an entirely different way. This vendor may never grow to the size of Oracle or SAP, but it is still a valid business model. Likewise, for some customers, better, faster, cheaper is well, better, faster, and cheaper. Josh is also right too. There are new application possibilities that can leverage processes across multiple enterprises and aggregate data from those enterprises in new and interesting ways.
That is not to say that one approach may not have a larger impact on software markets overall. Multi-tenant SaaS may be a discontinuous innovation in the Clayton Christensen sense and may continue to disrupt the power structures of many business application markets. However, it is not the only viable delivery model. All of the start up vendors I have seen for the past several years have had multi-tenant SaaS as their delivery model. I am waiting, and looking forward, to see some vendors with a different value proposition that provides a new and different spins on delivery models because of the customer business problem they are trying to solve.
