Group Search Strategy

Abstract

In this paper, we consider the fixed-sample and sequential strategies in group search, and we break the conventional wisdom that the sequential search strategy always dominates the fixed-sample strategy. We model group search as a game among group members, and we show that the fixed-sample strategy is preferable to the sequential strategy when the unit search cost (relative to the dispersion of the product value distribution) is very low or high enough. The sequential strategy has the information advantage compared to the fixed-sample strategy because decision makers can make use of information gained during the search process. However, due to the divergence of preferences in group search, the information advantage of the sequential strategy is reduced, whereas the commitment advantage of the fixed-sample strategy emerges: when the unit search cost is large, the fixed-sample strategy commits to a smaller number to search to save search cost, whereas when the unit search cost is small, the fixed-sample strategy chooses a larger number to search than the sequential strategy because it allows group members to commit to a number to search and prevents over-search upfront. Further, we show that our result is robust to a change in the distribution assumption or in the size of the group.