Essays in new product management

This dissertation examines issues related to new product management, a process that transforms market opportunities into new products for customers. Among the broad issues involved in new product management, we devote our attention to the management of generational product replacements, the planning of production ramp-up with the consideration of yield improvement through learning, and the channel's incentives in sharing unproved demand information.; Facing accelerated changes in technology and consumer taste, firms are under pressure to replace their product offerings swiftly. Mismanaged product offerings result in catastrophic failures. In Chapter 2, we characterize a decision framework by which a firm can manage generational product replacements under stochastic technological changes. First, we characterize an optimal threshold-based product replacement policy that maximizes the firm's expected total profit for a finite planning horizon. With this policy, the firm should replace its current product when the performance gap of the product is above a threshold; upon each product replacement, the firm should adopt the latest technology for the new product. Second, using stochastic ordering concepts, we quantify the negative impact from the accelerated technological changes on the expected total profit. Based on our results, we recommend that, instead of simply raising the frequency of its product replacement to chase technological trends, the firm put more emphasis on initiatives that can lower its product replacement cost.; When a new product is technologically ready, low production yield can prevent a firm from immediately ramping up the production of the product. The firm can improve its production yield through two types of learning: learning-before-doing and learning-by-doing. In Chapter 3, by modeling the yield improvement through both learning-before-doing and learning-by-doing, we develop a multi-stage model to analyze the optimal timing of production ramp-up. We show that, instead of evaluating every possible combination, a firm can achieve quite a robust outcome by simply comparing its expected profit from starting ramp-up now with that from starting ramp-up one period later, a myopic technique that is also known as the 1-Step Look-Ahead (1-SLA) policy. In particular, the 1-SLA policy is always optimal when learning-before-doing is more effective, or from another perspective, when the profit margin at production ramp-up is positive. Even when the 1-SLA policy is not optimal, it still can assist in screening out non-optimal policies. Numerical examples offer additional insights into production ramp-up time. For example, although production capacity is a partial substitute for production yield, it has limited impact on production ramp-up time.; Demand for a new product is highly uncertain, a risk factor that could lead to complete product failure. As the developer of a new product, the manufacturer in a manufacturer-retailer channel can reduce the uncertainty of the product's demand through observing progress in his product development project or receiving demand signals directly from customers. In Chapter 4, using the convex ordering concept, we first demonstrate that a centralized channel always benefits from improved demand information. Yet for a decentralized channel, the manufacturer's disclosure of his private demand information is essential to realize the benefit from reducing demand uncertainty. We show that the manufacturer's incentive to share his improved demand information depends on the supply contract that he signed with the retailer. Furthermore, mandating the manufacturer to disclose his improved demand information can actually reduce the total channel profit. We provide managerial insights by analyzing three widely used contract forms. We identify, for example, that the buyback contract always achieves information sharing while preserving channel performance.