What are we really good at? : product strategy with uncertain capabilities

Firms often learn about their own capabilities through their products' successes and failures. This paper explores the interaction between learning about capabilities and product strategy in a formal model. We consider a firm that can launch a sequence of products, where each product's success probability depends on the fit between the firm's capabilities and the product. A successful new product always causes the firm to become more optimistic about the capability most relevant for that product; however, it can also cause the firm to become less optimistic about some of its other capabilities, including capabilities the new product does not use. The firm's optimal forward-looking product strategy accounts for short-run expected profits as well as for the information value of learning for future decisions. We find that a product sharing few or even no capabilities with potential future products can have a greater information value than a product that shares more capabilities with future products and that learning about capabilities can affect the optimal sequence of product launch decisions