The uneven popularity of titles within a VOD content library, as expressed by its content demand profile, opens the door to improved efficiencies from multi-tiered storage architectures. Quantitative methods are shown for addressing the design and provisioning of such architectures. A single constant ‘r’ can describe the content demand profile. Analysis using ‘r’, along with storage media pricing ratios ’p’ and system streaming capacity ‘S’ results in a model for optimizing and balancing the mix of storage across multiple storage tiers. This analysis is applied to both 2-tier and 3-tier architectures, and is extended to consider the effects of future trends in content profiles, streaming demand, and storage media.