A multi-attribute service portfolio design problem

Abstract

Increasing popularity of cloud-based services has led to the emergence of cloud marketplaces where services from different providers are offered, usually in the form of a catalog. The customers' decision about buying offered services is based on idiosyncratic preferences regarding non-functional service attributes, e.g., price, provider reputation, and quality of service. Customer preferences are typically unknown to providers at the time the service portfolio (i.e. quality and price choices) is specified. Thus, from a microeconomic perspective, we have to deal with information asymmetry in markets, which complicates the challenge of finding the profit maximizing service portfolio. This paper presents a generic economic framework based on customer self-selection to address the above-mentioned optimization for a cloud service provider. The contribution is twofold: We characterize a multi-attributive customer preference function for cloud services based on a continuum of potential customers. Thereby each infinitesimal demand of a customer is characterized by a vector of minimum quality values for each of the different attributes and a maximum willingness to pay. The demand framework addresses the phenomenon of product cannibalization. We then formulate the service providers' optimal service portfolio design. This grants the provider maximal profit through optimal combination of potential values of the chosen attributes. © 2011 IEEE.