Customers' Preferences for Infrastructure Cloud Services

The pay-as-you-go pricing policies currently applied in Cloud Computing domain benefit customers with volatile demand but can lead to unmanageable utilization patterns for Cloud providers. Revenue management models that account for customer behavior enable to classify Cloud service users into customer segments and to increase the revenue for providers by offering different services. These theoretic models, however, make strong assumptions regarding the customer modeling. The results of our empirical study analyze, if revenue management models can be applied to Cloud Computing, and explain users' preferences, which has to be considered for the design of Cloud services. We show that users are willing to accept restrictions in order to receive discounts. The conjoint analysis reveals that operating system, price and support level of an IaaS offer have a major impact in the customers' selection process.