Demand estimation under multi-store multi-product substitution in high density traditional retail

M. Wan, Y. Huang, L. Zhao, T. Deng, J.C. Fransoo

Research output: Contribution to journalArticleAcademicpeer-review

5 Citations (Scopus)
1 Downloads (Pure)

Abstract

In large cities in emerging economies, traditional retail is present in a very high density, with multiple independently owned small stores in each city block. Consequently, when faced with a stockout, consumers may not only substitute with a different product in the same store, but also switch to a neighboring store. Suppliers may take advantage of this behavior by strategically supplying these stores in a coherent manner. We study this problem using consumer choice models. We build two consumer choice models for this consumer behavior. First, we build a Nested Logit model for the consumer choice process, where the consumer chooses the store on the first level and selects the product on the second level. Then, we consider an Exogenous Substitution model. In both models, a consumer may substitute at either the store level or the product level. Furthermore, we estimate the parameters of the two models using Markov chain Monte Carlo algorithm in a Bayesian manner. We numerically find that the Nested Logit model outperforms the Exogenous Substitution model in estimating substitution probabilities. Further, the information on consumers' purchase records helps improve the estimation accuracies of both the first-choice probabilities and the substitution probabilities when the beginning inventory level is low. Finally, we show that explicitly including such substitution behavior in the inventory optimization process can significantly increase the expected profit.
Original languageEnglish
Pages (from-to)99-111
JournalEuropean Journal of Operational Research
Volume266
Issue number1
DOIs
Publication statusPublished - 1 Apr 2018

Fingerprint

Substitution
Substitution reactions
Nested Models
Choice Models
Logit Model
Substitute
Consumer Behaviour
Markov Chain Monte Carlo Algorithms
Process Optimization
Model
Profit
Switch
Consumer behavior
Choose
Demand
Retail
Demand estimation
Product substitution
Markov processes
Profitability

Cite this

@article{029e19c8b149441ba822e97ab2f07e75,
title = "Demand estimation under multi-store multi-product substitution in high density traditional retail",
abstract = "In large cities in emerging economies, traditional retail is present in a very high density, with multiple independently owned small stores in each city block. Consequently, when faced with a stockout, consumers may not only substitute with a different product in the same store, but also switch to a neighboring store. Suppliers may take advantage of this behavior by strategically supplying these stores in a coherent manner. We study this problem using consumer choice models. We build two consumer choice models for this consumer behavior. First, we build a Nested Logit model for the consumer choice process, where the consumer chooses the store on the first level and selects the product on the second level. Then, we consider an Exogenous Substitution model. In both models, a consumer may substitute at either the store level or the product level. Furthermore, we estimate the parameters of the two models using Markov chain Monte Carlo algorithm in a Bayesian manner. We numerically find that the Nested Logit model outperforms the Exogenous Substitution model in estimating substitution probabilities. Further, the information on consumers' purchase records helps improve the estimation accuracies of both the first-choice probabilities and the substitution probabilities when the beginning inventory level is low. Finally, we show that explicitly including such substitution behavior in the inventory optimization process can significantly increase the expected profit.",
author = "M. Wan and Y. Huang and L. Zhao and T. Deng and J.C. Fransoo",
year = "2018",
month = "4",
day = "1",
doi = "10.1016/j.ejor.2017.09.014",
language = "English",
volume = "266",
pages = "99--111",
journal = "European Journal of Operational Research",
issn = "0377-2217",
publisher = "Elsevier",
number = "1",

}

Demand estimation under multi-store multi-product substitution in high density traditional retail. / Wan, M.; Huang, Y.; Zhao, L.; Deng, T.; Fransoo, J.C.

In: European Journal of Operational Research, Vol. 266, No. 1, 01.04.2018, p. 99-111.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Demand estimation under multi-store multi-product substitution in high density traditional retail

AU - Wan, M.

AU - Huang, Y.

AU - Zhao, L.

AU - Deng, T.

AU - Fransoo, J.C.

PY - 2018/4/1

Y1 - 2018/4/1

N2 - In large cities in emerging economies, traditional retail is present in a very high density, with multiple independently owned small stores in each city block. Consequently, when faced with a stockout, consumers may not only substitute with a different product in the same store, but also switch to a neighboring store. Suppliers may take advantage of this behavior by strategically supplying these stores in a coherent manner. We study this problem using consumer choice models. We build two consumer choice models for this consumer behavior. First, we build a Nested Logit model for the consumer choice process, where the consumer chooses the store on the first level and selects the product on the second level. Then, we consider an Exogenous Substitution model. In both models, a consumer may substitute at either the store level or the product level. Furthermore, we estimate the parameters of the two models using Markov chain Monte Carlo algorithm in a Bayesian manner. We numerically find that the Nested Logit model outperforms the Exogenous Substitution model in estimating substitution probabilities. Further, the information on consumers' purchase records helps improve the estimation accuracies of both the first-choice probabilities and the substitution probabilities when the beginning inventory level is low. Finally, we show that explicitly including such substitution behavior in the inventory optimization process can significantly increase the expected profit.

AB - In large cities in emerging economies, traditional retail is present in a very high density, with multiple independently owned small stores in each city block. Consequently, when faced with a stockout, consumers may not only substitute with a different product in the same store, but also switch to a neighboring store. Suppliers may take advantage of this behavior by strategically supplying these stores in a coherent manner. We study this problem using consumer choice models. We build two consumer choice models for this consumer behavior. First, we build a Nested Logit model for the consumer choice process, where the consumer chooses the store on the first level and selects the product on the second level. Then, we consider an Exogenous Substitution model. In both models, a consumer may substitute at either the store level or the product level. Furthermore, we estimate the parameters of the two models using Markov chain Monte Carlo algorithm in a Bayesian manner. We numerically find that the Nested Logit model outperforms the Exogenous Substitution model in estimating substitution probabilities. Further, the information on consumers' purchase records helps improve the estimation accuracies of both the first-choice probabilities and the substitution probabilities when the beginning inventory level is low. Finally, we show that explicitly including such substitution behavior in the inventory optimization process can significantly increase the expected profit.

U2 - 10.1016/j.ejor.2017.09.014

DO - 10.1016/j.ejor.2017.09.014

M3 - Article

VL - 266

SP - 99

EP - 111

JO - European Journal of Operational Research

JF - European Journal of Operational Research

SN - 0377-2217

IS - 1

ER -