This paper aims to propose a statistical method to measure the impacts of stockouts on demand, using a segmented linear regression model.
The proposed method is applied to data sets from large retail chains to measure the impacts of stockouts of an item on substitute items. The measured impacts of stockouts can be used to estimate the true demand of the sold-out item by recovering the lost demand (turned-away demand), as well as to estimate the true demand of the substitute item by reducing the extra demand.
This study found that estimated true demand by the proposed method improves sales forecasting and calculation of the annual expected revenue.
A new method to measure the impacts of stockouts on the demand of substitute items was proposed. The proposed method is practical, in that, it is conceptually simple, computationally efficient and applicable in general scenarios. Also, the proposed method is scalable for larger data sets.
