Search

Inaccurate forecasts of demand during promotions diminish the already meager profit margins of retailers. No forecasting method described in the literature can accurately account for the combination of seasonal sales variations and promotion-induced sales peaks over forecasting horizons of several weeks or months. We address this research gap by developing a forecasting method for seasonal, frequently promoted products that generates accurate predictions, can handle a large number of sales series, and requires minimal training data. In our method's first stage, we forecast the seasonal sales cycle by fitting a harmonic regression model to a decomposed training set, which excludes promotional and holiday sales, and then extrapolate that model to a testing set. In the second stage, we integrate the resulting seasonal forecast into a multiplicative demand function that accounts for consumer stockpiling and captures promotional and holiday sales uplifts. The final model is then fitted using ridge regression. We use sales data from a grocery retailing chain to compare the forecasting accuracy of our method with popular seasonal and promotion demand forecasting models at multiple aggregation levels for both short and long forecasting horizons. The significantly more accurate forecasts generated by our model attest to the merit of the approach developed here.

It is standard procedure to use short-term measures for designing customized marketing campaigns and evaluating their success. We consider the integration of customer lifetime value (CLV) thinking into the category selection problem for targeted coupons. This problem involves two primary considerations: should the coupon be a category-specific one? – and, if so, should short-term or rather long-term objectives govern the choice of category? Our paper is the first to address these questions by analyzing the effects of different customized targeting strategies for category coupons on short-term marketing costs and long-term CLV using an analytical model. We develop these targeting strategies with the help of a data-driven customer segmentation framework based on the dimensions of churn, frequency, and loyalty. Using data from a German hypermarket chain, we identify four refined targeting strategies that enable multi-category retailers to exploit the full potential of customized marketing by optimizing the trade-off between marketing cost and CLV considerations. Our analysis reveals that the targeting strategy most likely to be employed by retail practice – as compared with the one most commonly used – can yield a substantially higher CLV even as it reduces wasteful marketing costs.

Natural disasters, pandemics, and political nationalism force companies toward more responsive, flexible, and resilient assembly systems. For manufacturers, adaptability of the assembly process and...

The operations‐finance interface (OFI) jointly optimizes material, monetary, and information flows under intricate sources of uncertainty. To sketch the broad landscape in this emerging and interdisciplinary field, this study synthesizes literature across diverse themes and dispersed methodologies, screens systematically the research progression from original foundations to recent contributions in each major research stream, and thereby advocates future research innovations on prospective topics in light of the interconnections and potential reciprocity between operations and finance from risk management aspects.

In turbulent times featuring increased customisation, higher demand volatility, and shortened product life cycles, companies gain a competitive advantage by adopting a single yet highly flexible as...

In this case study, students combine data-based insights with strategic considerations to make fundamental business decisions at the German grocery retail chain Real. In response to dwindling numbers of customers and reduced revenues, Real developed the RealPro customer benefits program to achieve a quick turnaround. For a fixed annual fee, RealPro members receive substantial and permanent discounts of 20% on nonpromoted items from a broad range of food categories. Students employ data analytics methods to extract insights from the provided data set, which contains point-of-sale information from the actual market test of RealPro. Based on these insights, decisions concerning the rollout and design of the RealPro program must be made. We provide data analysis solutions in both Excel and R to analyze 75 thousand customer transactions. In the case extension, students can apply the difference-in-differences method and two covariate balancing algorithms for in-depth statistical analyses. For this purpose, we provide an additional unbalanced data set with 83 thousand transactions, on which the students can test and analyze propensity score matching and entropy balancing models.