I formulate an econometric model of consumer learning and experimentation about new products in markets for packaged goods that nests alternative sources of dynamics. The model is estimated on household level scanner data of laundry detergent purchases, and the results suggest that consumers have very similar expectations of their match value with new products before consumption experience with the good, but once consumers have learned their true match values they are very heterogeneous. I demonstrate that resolving consumer uncertainty about the new products increases market shares by 24 to 58%. The estimation results also suggest significant switching costs: removing switching costs increases new product market shares by 12 to 23%. Using counterfactual computations derived from the estimates of the structural demand model, I demonstrate that the presence of switching costs with learning changes the implications of the standard empirical learning model: the intermediate run impact of an introductory price cut on a new product’s market share is significantly greater when the only source of dynamics is switching costs as opposed to when both learning and switching costs are present, which suggests that firms should combine price cuts with introductory advertising or free samples to increase their impact. Because my model includes two different types of dynamics, I am able to assess the impact of ignoring learning or switching costs on the model’s imputed long run price elasticities by reestimating the model assuming that one of these dynamics is not present. I find that ignoring learning will i) lead to underestimates of the own price elasticities of new products by 30%, ii) will underestimate the cross-price elasticities between new and established products by up to 90%, iii) will overestimate the cross-price elasticities of established products by up to 15%. Ignoring switching costs will lead to underestimates of own price elasticities of up to 60%, and underestimates of crossprice elasticities of up to 90%.