Recently cloud computing gains more and more attention from academia as well as practitioners. Most of the research so far focused on its technical issues, while marketing issues of cloud computing have barely been discussed in literature, especially customer considerations and pricing. This study intends to fill this gap and approaches cloud computing from a customer perspective. A survey was conducted on small and medium enterprises in Singapore . By using a choice based conjoint analysis we provide a valuable insight into customers' preferences on the design of cloud computing. In contrast to other studies, relative importance of cloud service attributes is identified and shows that economic reasons are not the most important issues as expected before. Moreover cluster analysis results show preference heterogeneity. As a consequence, three different customer segments are defined which should be provided with different cloud services. The customer heterogeneity reveals possibilities for second price discrimination based on versioning. Moreover, the study shows the existence of flat rate and pay per use tariff biases within cloud computing. The customer bias is significantly influenced by flexibility and insurance effects. Therefore the selection of offered tariffs should not only be based on economic considerations, but take also customer preferences into account.