Retail Promotion Forecasting: A Comparison of Modern Approaches

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

Abstract

Promotions at retailers are an effective marketing instrument, driving customers to stores, but their demand is particularly challenging to forecast due to limited historical data. Previous studies have proposed and evaluated different promotion forecasting methods at product level, such as linear regression methods and random trees. However, there is a lack of unified overview of the performance of the different methods due to differences in modeling choices and evaluation conditions across the literature. This paper adds to the methods the class of emerging techniques, based on ensembles of decision trees, and provides a comprehensive comparison of different methods on data from a Danish discount grocery chain for forecasting chain-level daily product demand during promotions with a four-week horizon. The evaluation shows that ensembles of decision trees are more accurate than methods such as penalized linear regression and regression trees, and that the ensembles of decision trees benefit from pooling and feature engineering.

Original languageEnglish
Title of host publicationAdvances in Production Management Systems. Towards Smart Production Management Systems : IFIP WG 5.7 International Conference, APMS 2019, Austin, TX, USA, September 1–5, 2019, Proceedings, Part II
Volume2
PublisherSpringer
Publication date2019
Pages575-582
ISBN (Print)978-3-030-29995-8
ISBN (Electronic)978-3-030-29996-5
DOIs
Publication statusPublished - 2019
EventIFIP WG 5.7 International Conference, APMS 2019 - Austin, United States
Duration: 1 Sep 20195 Sep 2019

Conference

ConferenceIFIP WG 5.7 International Conference, APMS 2019
CountryUnited States
CityAustin
Period01/09/201905/09/2019
SeriesIFIP AICT - Advances in Information and Communication technology
Volume567
ISSN1868-4238

Fingerprint

Retail Promotions
Decision tree
Evaluation
Linear regression
Discount
Regression method
Grocery
Retailers
Marketing
Regression tree
Forecasting method
Choice modelling
Pooling

Cite this

Bojer, C. S., Dukovska-Popovska, I., Christensen, F. M. M., & Steger-Jensen, K. (2019). Retail Promotion Forecasting: A Comparison of Modern Approaches. In Advances in Production Management Systems. Towards Smart Production Management Systems: IFIP WG 5.7 International Conference, APMS 2019, Austin, TX, USA, September 1–5, 2019, Proceedings, Part II (Vol. 2, pp. 575-582). Springer. IFIP AICT - Advances in Information and Communication technology, Vol.. 567 https://doi.org/10.1007/978-3-030-29996-5_66
Bojer, Casper Solheim ; Dukovska-Popovska, Iskra ; Christensen, Flemming Max Møller ; Steger-Jensen, Kenn. / Retail Promotion Forecasting : A Comparison of Modern Approaches. Advances in Production Management Systems. Towards Smart Production Management Systems: IFIP WG 5.7 International Conference, APMS 2019, Austin, TX, USA, September 1–5, 2019, Proceedings, Part II. Vol. 2 Springer, 2019. pp. 575-582 (IFIP AICT - Advances in Information and Communication technology, Vol. 567).
@inproceedings{ef7251f73b3447eca5b0f7ceb79d9c64,
title = "Retail Promotion Forecasting: A Comparison of Modern Approaches",
abstract = "Promotions at retailers are an effective marketing instrument, driving customers to stores, but their demand is particularly challenging to forecast due to limited historical data. Previous studies have proposed and evaluated different promotion forecasting methods at product level, such as linear regression methods and random trees. However, there is a lack of unified overview of the performance of the different methods due to differences in modeling choices and evaluation conditions across the literature. This paper adds to the methods the class of emerging techniques, based on ensembles of decision trees, and provides a comprehensive comparison of different methods on data from a Danish discount grocery chain for forecasting chain-level daily product demand during promotions with a four-week horizon. The evaluation shows that ensembles of decision trees are more accurate than methods such as penalized linear regression and regression trees, and that the ensembles of decision trees benefit from pooling and feature engineering.",
author = "Bojer, {Casper Solheim} and Iskra Dukovska-Popovska and Christensen, {Flemming Max M{\o}ller} and Kenn Steger-Jensen",
year = "2019",
doi = "10.1007/978-3-030-29996-5_66",
language = "English",
isbn = "978-3-030-29995-8",
volume = "2",
series = "IFIP AICT - Advances in Information and Communication technology",
publisher = "Springer",
pages = "575--582",
booktitle = "Advances in Production Management Systems. Towards Smart Production Management Systems",
address = "Germany",

}

Bojer, CS, Dukovska-Popovska, I, Christensen, FMM & Steger-Jensen, K 2019, Retail Promotion Forecasting: A Comparison of Modern Approaches. in Advances in Production Management Systems. Towards Smart Production Management Systems: IFIP WG 5.7 International Conference, APMS 2019, Austin, TX, USA, September 1–5, 2019, Proceedings, Part II. vol. 2, Springer, IFIP AICT - Advances in Information and Communication technology, vol. 567, pp. 575-582, IFIP WG 5.7 International Conference, APMS 2019, Austin, United States, 01/09/2019. https://doi.org/10.1007/978-3-030-29996-5_66

Retail Promotion Forecasting : A Comparison of Modern Approaches. / Bojer, Casper Solheim; Dukovska-Popovska, Iskra; Christensen, Flemming Max Møller; Steger-Jensen, Kenn.

Advances in Production Management Systems. Towards Smart Production Management Systems: IFIP WG 5.7 International Conference, APMS 2019, Austin, TX, USA, September 1–5, 2019, Proceedings, Part II. Vol. 2 Springer, 2019. p. 575-582 (IFIP AICT - Advances in Information and Communication technology, Vol. 567).

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Retail Promotion Forecasting

T2 - A Comparison of Modern Approaches

AU - Bojer, Casper Solheim

AU - Dukovska-Popovska, Iskra

AU - Christensen, Flemming Max Møller

AU - Steger-Jensen, Kenn

PY - 2019

Y1 - 2019

N2 - Promotions at retailers are an effective marketing instrument, driving customers to stores, but their demand is particularly challenging to forecast due to limited historical data. Previous studies have proposed and evaluated different promotion forecasting methods at product level, such as linear regression methods and random trees. However, there is a lack of unified overview of the performance of the different methods due to differences in modeling choices and evaluation conditions across the literature. This paper adds to the methods the class of emerging techniques, based on ensembles of decision trees, and provides a comprehensive comparison of different methods on data from a Danish discount grocery chain for forecasting chain-level daily product demand during promotions with a four-week horizon. The evaluation shows that ensembles of decision trees are more accurate than methods such as penalized linear regression and regression trees, and that the ensembles of decision trees benefit from pooling and feature engineering.

AB - Promotions at retailers are an effective marketing instrument, driving customers to stores, but their demand is particularly challenging to forecast due to limited historical data. Previous studies have proposed and evaluated different promotion forecasting methods at product level, such as linear regression methods and random trees. However, there is a lack of unified overview of the performance of the different methods due to differences in modeling choices and evaluation conditions across the literature. This paper adds to the methods the class of emerging techniques, based on ensembles of decision trees, and provides a comprehensive comparison of different methods on data from a Danish discount grocery chain for forecasting chain-level daily product demand during promotions with a four-week horizon. The evaluation shows that ensembles of decision trees are more accurate than methods such as penalized linear regression and regression trees, and that the ensembles of decision trees benefit from pooling and feature engineering.

U2 - 10.1007/978-3-030-29996-5_66

DO - 10.1007/978-3-030-29996-5_66

M3 - Article in proceeding

SN - 978-3-030-29995-8

VL - 2

T3 - IFIP AICT - Advances in Information and Communication technology

SP - 575

EP - 582

BT - Advances in Production Management Systems. Towards Smart Production Management Systems

PB - Springer

ER -

Bojer CS, Dukovska-Popovska I, Christensen FMM, Steger-Jensen K. Retail Promotion Forecasting: A Comparison of Modern Approaches. In Advances in Production Management Systems. Towards Smart Production Management Systems: IFIP WG 5.7 International Conference, APMS 2019, Austin, TX, USA, September 1–5, 2019, Proceedings, Part II. Vol. 2. Springer. 2019. p. 575-582. (IFIP AICT - Advances in Information and Communication technology, Vol. 567). https://doi.org/10.1007/978-3-030-29996-5_66