The use of disaggregated demand information to improve forecasts and stock allocation during sales promotions: A simulation and optimisation study using supermarket loyalty card data

Sheraz Alam Malik, Andrew Fearne, Jesse O Hanley

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)
11 Downloads (Pure)


Our work highlights the importance of using disaggregated demand information at store level to improve sales forecasts and stock allocation during sales promotions. Monte Carlo simulation and optimisation modelling were used to estimate short-term promotional impacts. Supermarket loyalty card data was used from a major UK retailer to identify the benefits of using disaggregated demand data for improved forecasting and stock allocation. The results suggest that there is a high degree of heterogeneity in demand at individual store level due to number of factors including the weather, the characteristics of shoppers, the characteristics of products and store format, all of which conspire to generate significant variation in promotional uplifts. The paper is the first to use supermarket loyalty card data to generate store level promotional forecasts and quantify the benefits of disaggregating the allocation of promotional stock to the level of individual stores rather than regional distribution centres.

Original languageEnglish
Pages (from-to)339-357
Number of pages19
JournalInternational Journal of Value Chain Management
Issue number4
Early online date17 Oct 2019
Publication statusPublished - 2019


  • Demand forecasting
  • Monte Carlo simulation
  • Optimisation
  • sales promotions
  • Stock allocation
  • Supermarket loyalty card data

Cite this