Dynamic quality decisions of software-as-a-service providers based on customer perception

•A unique optimal quality policy and the optimal initial quality exist for the SaaS provider.•The SaaS provider has a constrained opportunistic behavior, which makes the quality decisions transiently fluctuating.•Customers’ expectations on quality improvement work positively on a SaaS provider’s dynamic quality decisions.•The value of dynamic pricing adjustment is limited, but it is high when choosing an inappropriate initial actual quality.•The SaaS provider should set a high initial quality when customer perception is uncertain.

Software-as-a-service (SaaS) allows customers to use software over the Internet by paying a subscription fee rather than by buying shrink-wrapped software and installing it on their computers. To maintain customer interests in SaaS, the provider’s dynamic quality decision is important. Thus, we consider a dynamic programming model that assumes demand is quality sensitive and influenced by customer perception. In both finite and infinite planning horizons, we show that a unique optimal policy exists for the SaaS provider to set quality periodically. We demonstrate that the SaaS provider may have a constrained opportunistic behavior towards its quality decisions when considering customer perception. This behavior results in a transient fluctuating quality decision path before it converges to a steady state. We find that the software’s initial quality plays a critical role in long term decisions, and that customers’ expectations of quality improvement positively affect the quality, although the actual improvement may not meet their expectations.