Work

Valikoimasta johtuvan substituution mallintaminen vähittäiskaupassa

Year: 2019
Type: dissertation
Abstract: Deciding which products to include in the assortment of a store is one of the most basic decisions a retailer has to make, but finding out which assortment produces the highest profits or margins is a difficult problem. Failing to offer the optimal products causes loss of profits or even loss of customers for the retailer. When customers are looking for a specific product but find that it is unavailable, they may choose to buy another similar product instead. This is called substitution, and the amount of substitution affects the demand of products. Therefore, the optimal assortment also depends on the amount of substitution, and hence, many assortment planning solutions incorporate the effect of substitution in them. However, only one study was found where substitution has been estimated directly. The goal of this thesis is to find how assortment-based substitution in retail can be directly estimated. In this thesis the methodology for estimating assortment-based substitution is presented and then applied to a retail receipt data set. The results indicate that the accuracy of the substitution estimation method depends heavily on the accuracy of the demand forecasts that are used as inputs for the substitution estimation. The forecasting models used in this thesis are not able to predict demand accurately for slow-moving products, which limits the accuracy of the substitution estimation method. Additionally, the results of this thesis show that the footfall of a store, i.e. how many customers visit a store on a given day, can be estimated very accurately.
Author: Vainio, Ville
Institution: Aalto University
Language: English
Open Access status: green
OpenAlex ID: W7114137297
Cites: 28
Cited by:
Related to: 9