Software as a Service (SaaS) has been a buzz word for a while. Does not seem to take off too well. I attribute this to two main reasons. Larger vendors are still trying to figure out how to move their perpetual license customers into this model (i.e. if they want) and trying to create a compelling business case to ensure that they do not cannibalize exiting revenue streams.
However from a financial perspective many have argued that SaaS has a very different financial impact on an IT organization. This is simply due to the fact that from a license plus annual maintenance model, the corporations suddenly start incurring an "annual fee". They would argue that this changes the impact of a SaaS based offering to their cash flow projections. In other words, the initial license fee is a Capital Expense (CAPEX) and the annual maintenance is an Operating Expense (OPEX). Therefore there will be a bigger drop on the cash flow projections for the initial year due to the license costs and then a fixed decrease due to the annual maintenance OPEX. In a SaaS based model there is no CAPEX and everything is a OPEX which leads to the fact that there is not CAPEX and hence no impact of depreciation. Apparently this kind of sounds right....Does it? The financial impact of the SaaS model is analogous to an "Operating Lease" when the books create a "fictitious asset" and "depreciate" it over the years. This is technically how it must be treated. Hence, CIOs must not worry about the impact of SaaS on cash flows as it is no different that other software purchases.
Looking at the operating aspects of Saas, is it any better than traditional software from a maintenance or operational perspective? Apparently the SaaS model seems to reduce OPEX. However, I think it does that marginally. The biggest merit of the model is its impact on operational RISK. One can argue that since it is a hosted solution (hosted at someone else' site), it adds to the risk since there is less control on the day-to-day functioning of the solution. Vendors hosting solutions are usually using a robust and proven infrastructure. They are also "very good" at what they do - host solutions. Finally they negotiate Service Level Agreements (SLAs) with their customers on the downtime of the hosted solution within acceptable levels of threshold of the customers' tolerance. Hence, I would say a SaaS model reduces operational risk. To turn the argument around, such a model has a higher opportunity cost - the cost of downtime if hosted within the company infrastructure. It may also be argued that a hosted solution is vulnerable to security intrusions. This increases the security risk of the solution which is a significant component of operational risk. The decision to use a hosted solution is predicated on a careful evalution of the risks that it exposes and mitigates. Accessing a SaaS model from an OPEX perspective is not completely apporpriate.
From a SaaS provider's perspective, such a model is ideal to leverage Economies of Scale once the number of customers reach a certain number. SaaS vendors usually price their offering as per seat, per usage or on subscription basis. In a perfect competition scenaro, the SaaS vendor will maximize his profits when his Margical Cost is equal to his Marginal Revenue. However, when economies of scale are achieved, Marginal Cost tends to be minimal or equal to zero. At that point, if the vendor can differentiate his offering from his competitors, he can transition to a Price Discrimination model to price his customers differently. This leads to higher profits over a period of time. However, the key message here consists of two variables - large number of customers to achieve economies of scale and differentiated service offering to leverage a Price Discrimination strategy.
A SaaS model seems to work very well for applications that are independent in nature i.e. not tightly integrated with the other enterprise applications. SaaS based solutions work really well if their dependency on other applications is low or predictable over a time period. Hence CRM systems that are not tightly integrated with Order Management or Provisioning systems seem to be an excellent are where this model seems to be getting a lot of traction and success - a.k.a. SalesForce and NetSuite. However, if these are tightly integrated with other ERP systems the efficacy reduces leading to a non-economic solution offering. Other areas that seem to benefit from a SaaS model are industries that require "Information on Demand" - such as Financial Services Analytics, Video on Demand, News etc.
So if all is nice and dandy with SaaS, why is it not picking-up or slow to pick-up? The answer is simple - the diffusion of this technology into the "Late Majority" has not yet happened. This is due to two reasons. Since the profitability of a SaaS solution requires a larger gestation period, vendors are reluctant to offer it as a primary delivery channel of their products. This is also due to the aggressive competition. Competitive forces are not allowing software vendors to attain the desired economies of scale leading to revert back to a licensing model where they make money from the maintenance contracts. From the customers' perspective the perception of risk offsets the benefits of the model.
Who says "Customers are rational"?
