ABSTRACT An emerging trend in multimedia applications such as distance education and corporate training is to service clients according to their convenience in terms of start time of the multimedia content and the quality of reception. Examples include clients requesting for a movie to start at a convenient time specified by (t+di) where t is the current time and di is the delay tolerance acceptable to client i. Such applications typically involve a Closed User Group network that exhibits heterogeneous characteristics, where a Content Service Provider (CSP) services requests from geographically dispersed clients synchronously. It is important for a CSP to utilize resources such as buffers, transcoders, and caches judiciously in order to minimize costs while serving clients with their required quality in order to maximize revenues. We approach the problem in the following steps: (i) Determine the optimal quality deliverable to the clients while satisfying their delay tolerance assuming static network characteristics for the duration of the play out (ii) Since the playback need to start at the requested time, determine the optimal placement of buffers, caches, and transcoders such that resource utilization is maximized and client device constraints are satisfied (iii) Use admission control and scheduling to consider the trade off in revenue if clients were admitted (dynamic arrivals) while satisfying the admitted clients’ requirements by maximizing resource utilization. We have developed an optimization-based approach to determine the best quality that can be delivered to the clients using resources such as buffers and transcoders. Simulation results demonstrate the usefulness of exploiting client delay tolerance specifications for delivering enhanced Quality of Service (QoS) with little or no additional resources.