157 /304
157
臺大管理論叢
第
28
卷第
1
期
percentages of purchase decisions in the condition of congruence are very similar to that
in the baseline (Quality: Broad-appeal (.77) vs. Narrow-appeal (.23); Social: Broad-appeal
(.78) vs. Narrow-appeal (.22); Baseline: Broad-appeal (.78) vs. Narrow-appeal (.21). This
suggests that the effect of popularity information and inference do not impact consumers’
purchase decisions so much. However, participants’ average percentages of purchase
decisions in the condition of incongruence are very different from that in baseline and the
patterns between the two inferences are dramatically different. This suggests that the
effect of popularity information and inferences jointly influence participants’ purchase
decisions, which matches the central belief of this research.
5.2 Conclusions
The emergence of the Internet has fundamentally changed how sellers and consumers
supply and respond to popularity information. Specifically, Internet technology has
enabled e-businesses to carry a wide assortment of products, including both broad-appeal
and narrow-appeal products and allows them to implement up-to-the-minute sales volume
next to every product with relatively low cost and effort (Brynjolfsson et al., 2010).
Moreover, due to a wide variety of product selections on the Internet, products in similar
categories can be easily placed together for ease of access. This suggests that consumers
may often see broad-appeal and narrow-appeal products in the same category placed
together, each with their own higher or lower sales volume, which matches the conditions
of congruence and incongruence described in this research. All of the above have
demonstrated how the information of sales volume and breadth of appeal are linked and
may jointly influence consumers’ purchase decision in e-commerce. Therefore, we have
intentionally developed the fundamental setting of our research to match with the unique
characteristics on the Internet in order to examine the joint effect in popularity
information. In addition, we have also included different inferences (quality evaluation vs.
social comparison) into the research model to examine its possible moderating effect. We
hope the empirical findings of this research have shed additional light in understanding the
complicated effect of popularity information on the Internet and provide meaningful
theoretical and managerial implications.