When and how is the Internet likely to decrease price competition?

Conventional wisdom seems to claim that, by lowering the cost of distributi on and by making search easier for consumer, the introduction of the Intern et is likely to intensify price competition. This paper intends to challeng e this view by asking: When and how is the Internet likely to decrease the level of price competition between firms? To answer this question, we devel op an analytic model with the following characteristics. On the demand side , consumers need to gather information on two types of product attributes: digital attributes (which can be communicated on the Web at very low cost) and nondigital attributes (for which physical inspection of the product is necessary). Consumers choose between two brands but are familiar with the n ondigital attributes of only the brand purchased on the last purchase occas ion. On the supply side, firms use traditional stores and the Internet to i nform consumers about their products' attributes and to sell their products . In this setup, we show that the impact of the Internet on competition wil l be radically different depending on the relative importance of parameters describing the relevant shopping and distribution context. Specifically, w e find that the introduction of the Internet might lead to monopoly pricing when (1) the proportion of Internet users is high enough, (2) when nondigi tal attributes are relevant but not overwhelming, (3) when consumers have a more favorable prior about the brand they currently own, and (4) when the purchase situation can be characterized by "destination shopping". More sur prising, we also show that in such cases, the use of the Internet not only leads to higher prices but can also discourage consumers from engaging in s earch. As such, an important message of the paper is that under some condit ions the Internet might represent an opportunity for firms to leverage thei r brand loyalty and increase their profits. The intuition behind our results is the following. The Internet allows cons umers to evaluate digital attributes easily, i.e., without visiting the sto res. However, nondigital attributes can only be evaluated through physical presence. As such, for goods where both types of attributes are important, the introduction of the Net changes the effective cost of search for consum ers. Without the Internet the cost of search is the cost of visiting more t han one store. With the introduction of the Net however, nonsearching consu mers do not have to undertake the shopping trip at all because they can ord er products on the Net. Thus, in the presence of the Internet the cost of s earch is related to the cost of undertaking the entire shopping trip. In th e case of destination shopping (i.e. when the fixed cost of undertaking the shopping trip is higher than the cost of visiting an additional store), th e presence of the Internet creates higher effective search costs for consum ers. Given this shift of paradigm in search costs due to the Internet, cons umers may not take the risk of searching for products with better nondigita l attributes, but instead, remain with the product they are familiar with. This results in increased consumer loyalty, which induces firms to increase their prices. Our results have important managerial implications. First, they provide gui delines for firms on when (i.e. for which product categories) they should c onsider expanding their distribution network to the Internet. In this respe ct, an important additional insight of the paper is that the Internet can l ower price competition and lead to reduced consumer search even if it is mo re expensive than the traditional distribution channel. This can easily be the case if distribution through the Internet represents additional costs s uch as the costs associated with shipping and handling and return policies. Second, the paper also provides guidelines on how to plan the firm's Inter net strategy. Interestingly, the results suggest that with the general avai lability of the Internet the role of stores might actually become more impo rtant. While we do not explicitly model a dynamic market, our findings toge ther with Klemperer's (1987) results suggest that stores might have a key r ole in consumer acquisition, while the Internet can help leverage the acqui red customer base through demand fulfillment. This might imply that for cer tain product categories, firms should actually allocate additional resource s to improve their in-store environment when considering the Internet as a complementary distribution channel.