group buying strategies available through alphatise

becoming more and more popular in e-commerce. In this paper, we derive the utility consumers obtain from GB by recognizing both positive network effect and negative network effect buying with others. According to their valuations for the product and attitudes to network effects, consumers are classified into two segments. From the perspective of sellers, we compare three possible strategies, i.e., offering only IB option, offering only GB option, and offering both options, and derive the optimal decisions on price and quantity for each strategy.

Our result shows that both the network effect and consumer heterogeneity play a significant role in choosing optimal strategy. In particular, offering only GB dominates offering only IB when the positive network effect is sufficiently high or the proportion of low valuation segment is relatively large. We also find that offering both options simultaneously can improve the seller’s revenue compared to offering only IB option, while the seller has to decide offering only GB or offering GB along with IB based on actual market situations.

Several managerial implications can be derived from our study. The positive network effect is quite significant for first-time buyers. Based on this fact, the firm selling a new product can enhance consumers’ valuation and expand the market quickly by offering only GB option. When the product becomes an “old” one, more consumers identify highly with the product and the negative network effect is more significant.

In this case, the firm should choose IB strategy (or MIX strategy offering IB and GB simultaneously) rather than GB strategy. Some limitations apply to our model. First, this paper only considers the decision-making problem for a monopoly seller. It is of interest to consider the case under a competitive environment. Second, instead of uniform distribution for consumers’ valuation, one can consider a general distribution and examine the influence of distribution variation on equilibrium outcomes. Finally, one can also extend the analysis to the cases with random demand or/and imperfect information. However, it might be outrageously complex and challenging.