Using machine learning to predict the next purchase date for an individual retail customer

Droomer, Marli ; Bekker, James (2020-11-11)

CITATION: Droomer, M. & Bekker, J. 2020. Using machine learning to predict the next purchase date for an individual retail customer. South African Journal of Industrial Engineering, 31(3):69-82, doi:10.7166/31-3-2419.

The original publication is available at http://sajie.journals.ac.za

Article

ENGLISH ABSTRACT: Targeted marketing has become more popular over the last few years, and knowing when a customer will require a product can be of enormous value to a company. However, predicting this is a difficult task. This paper reports on a study that investigates predicting when a customer will buy fast-moving retail products, by using machine learning techniques. This is done by analysing the purchase history of a customer at participating retailers. These predictions will be used to personalise discount offers to customers when they are about to purchase items. Such offers will be delivered on the mobile devices of participating customers and, ultimately, physical, general paper-based marketing will be reduced.

AFRIKAANSE OPSOMMING:Teiken-bemarking het in die laaste jare gewild geword en dit kan baie waardevol wees vir ʼn onderneming om te weet wanneer ʼn kliënt ʼn produk benodig. Hierdie soort voorspelling is egter baie moeilik, en hierdie artikel beskryf ʼn studie wat met behulp van masjienleer-tegnieke ondersoek het wanneer ʼn kliënt vinnig-bewegende kleinhandel produkte sal koop. Dit is gedoen deur die koopgeskiedenis van ʼn kliënt by deelnemende kleinhandelaars te ontleed. Hierdie voorspellings sal gebruik word om persoonlike afslagaanbiedinge aan kliënte te maak wanneer hulle produkte wil koop. Hierdie aanbiedinge sal op deelnemende kliënte se mobiele toestelle aangebied word en uiteindelik sal veralgemeende, papier-gebaseerde bemarking verminder word.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/108929
This item appears in the following collections: