Product bundling is a business strategy that packages (either physically or logically), prices and sells groups of two or more distinct products or services as a single economic entity. This practice exploits variations in the reservation prices and the valuations of a bundle vis-à-vis its constituents. Bundling is an effective instrument for price discrimination, and presents opportunities for enhancing revenue without increasing resource availability. However, optimal bundling strategies are generally difficult to derive due to constraints on resource availability, product valuation and pricing relationships, the consumer purchase process, and the rapid growth of the number of possible alternatives.

This dissertation investigates two different situations—vertically differentiated versus independently valued products—and develops two different approaches for revenue maximization opportunities using product bundling, when resource availability is limited. For the vertically differentiated market with two products, such as the TV market with prime time and non-prime time advertising, we derive optimal policies that dictate how the seller (that is, the broadcaster) can manage their limited advertising time inventories. We find that, unlike other markets, the revenue maximizing strategy may be to offer only the bundle, only the components, or various combinations of the bundle and the components. The optimality of these strategies critically depends on the availability of the two advertising time resources. We also show how the network should focus its programming quality improvement efforts, and investigate how the “value of bundling,” defined as the network’s and the advertisers’ benefit from bundling, changes as the resource availabilities change. We then propose and study a bundling model for the duopolistic situation, and extend the results from the monopolistic to the duopolistic case.

For the independently valued products, we develop stochastic mathematical programming models for pricing bundles of n components. Specializing this model for two components in a deterministic setting, we derive closed-form optimal product pricing policies when the demand functions are linear. Using the intuition garnered from these analytical results, we then investigate two procedures for solving large-scale problems: a greedy heuristic, and a decomposition method. We show the effectiveness of both methods through computational experiments.