158 /304
產品受歡迎資訊與網路購物：以消費者解讀為干擾變數
158
5.3 Theoretical Implications
Different from the traditional research on popularity information that was based
solely on one inference, this research has considered two possible inferences and
incorporated them with characteristics common on the Internet. The empirical findings of
this research show: When breadth of appeal and sales volume are congruent, consumers
have a higher tendency to purchase a broad-appeal over a narrow-appeal product.
Moreover, the average percentages of purchase decisions allocated to broad-appeal and
narrow-appeal choices are similar to that in the baseline, suggesting the effect of
popularity is not apparent in this condition. These results match the belief in Tucker and
Zhang’s (2011) research that popularity information does not positively affect consumers’
purchase decisions if its high sales volume is driven by its naturally wide appeal to the
mainstream market.
On the other hand, when breadth of appeal and sales volume are incongruent, results
with the quality inference show that the percentages of choosing a narrow-appeal product
are significantly higher than those in the condition of congruence, which also matches
Tucker and Zhang’s (2011) belief. Yet, this research has gone beyond Tucker and Zhang’s
(2011) research scope by incorporating the social inference into the research model and
making comparisons with the quality inference. Results with the social inference show
that the percentage of purchasing a narrow-appeal product in the condition of
incongruence is also significantly increased when compared to the condition of
congruence. However, the magnitude of increase is not as strong as when the quality
inference is dominant because some consumers choose the broad-appeal product in this
condition. These different findings from different dominant inferences may clarify the
mixed evidence in the literature that popularity information may boost the sales for a
broad-appeal product as described in the concept of “winner-takes-all” or a narrow-appeal
product as described in the concept of “long-tail” (Brynjolfsson et al., 2011; Brynjolfsson
et al., 2010). Our research provides legitimate explanations for these divergent situations
and allows the effect of popularity information to be more generally applied in the context
of online purchases.
Finally, much research in the Marketing literature has focused on understanding how
online social influences could be exerted by different roles and channels on the Internet
(Bickart and Schindler, 2001; Chevalier and Mayzlin, 2006; Zhu and Zhang, 2010). Our
research has furthermore made contributions in understanding how online social
comparison could be exerted by celebrities and general consumers through popularity