In this study, the problem environment consists of two fast fashion retailing firms where one can obtain the other's selling data from an outside agent. The aim of the study is to assess the value of such a third party and the each information piece obtained from it. This paper proposes a labelling algorithm based on a dynamic discrete choice modelling approach and uses it to study the effect of different attributes on estimation performance. Our extensive numerical study shows that the dynamic discrete choice-based procedure leads to highly accurate estimation for the hidden demand factors. Secondly, we show that if the seller is a strategic decision maker, it actually enhances the degree of accuracy as it becomes easier for the discrete-choice model to reverse-engineer the true values of the hidden demand factors. Finally, as the depth of discount increases, the degree of accuracy decreases. This comes from the fact that the deeper discounts disincentivizes the seller to offer discounts, which in turn reduces the chances of observing the demand at different price points.
Published March 2017 , 13 pages