A decision support approach to achieve competitive advantage for a hypermarket chain

This paper describes the decision support approach used in the development process of the S Group's Prisma hypermarket chain in Finland. The management was looking for a new and sustainable operating model for the rapidly growing chain, and contacted the author to consult in the process. Fierce competition forced the search for new business ideas, tools and methods that would provide a clear competitive advantage. To find new perspectives, we decided to use statistical approaches and various decision support system options, such as multi-criteria modelling. A database was available for research and analysis, including data on purchasing behavior and key performance indicators (KPI). The approach had to take into account the role and impact of customers. It was highly important to include customer behavior in the analysis using shopping basket data. Shopping basket data was central in the current paper. From these, an observation matrix was created combining shopping basket data, product data and customer background information. Using multivariate methods, customer groupings and profiles were created with the data from the observation matrix. Using the customer profile and KPI data, a multi-criteria decision support system was produced to support strategic planning. The decision support system (DSS) model was created together with a market chain operational expert and an external methodological expert. We used the VIG software package developed by Korhonen (Belg J Oper Res Stat Comput Sci 27(3):15, 1987) to solve the problem because it is easy to use and requires no prior knowledge of computers or multi-objective linear programming models. Pareto Race plays a central role in the VIG system. The chain expert easily learned how to use and work with the model. The results were immediately visible and could be used to examine alternatives and assess their appropriateness. It was decided to present five different scenarios to the hypermarket chain management. The main objective of the development process was to develop a strategy that would provide the Prisma hypermarket chain with a long-term competitive advantage. Various models were developed and used to support the strategy work by analysing and exploring the data collected, prioritising and selecting decision options. Two currently retired managers (Mönkkönen, S Group, the chain manager, Prisma chain, Interview 02.06.2021, 2021), who were involved in the development process, rated the strategy process as very successful and the modelling carried out during the process significantly supported decision-making. The immediate help of DSS modelling for decision making comes from being able to provide decision makers with reasonable, better solution options to support their decision making. The final impact of decisions could be evaluated after a longer period of time, which in the case of the Prisma development project results means several comparable financial years. Finland suffered exceptionally badly from the financial crisis and the global economic downturn in 2008–2009. The Prisma chain has survived the periods and crises described above without any loss-making years, and the whole chain has grown from 16 units in 1992 to 68 units in 2020.

Approach to Kirana Store Product Arrangement Using Machine Learning

Market Basket Analysis (MBA) is a method for determining the association between entities, and it has often been used to study the association between products in a shopping basket. Trained Computer vision models are able to recognize objects in photos so accurately that it can even outperform humans in some instances. This study shows that combining objective detection techniques with market basket analysis can assist Stores/Kirana in organizing the products effectively. With the use of MBA and Object detection, we formulated recommendations for store arrangements along with putting a recommendation engine on top to help shoppers. After deploying this to local Kirana stores, the Kirana store was able to see an increase of 7% in the sale. The recommendation engine performed better than just the domain knowledge of the kirana store.

EXPRESS: Households under Economic Change: How Micro- and Macroeconomic Conditions Shape Grocery Shopping Behavior

Economic conditions may significantly affect households' shopping behavior and, by extension, retailers' and manufacturers' firm performance. By explicitly distinguishing between two basic types of economic conditions—micro conditions in terms of households' personal income and macro conditions in terms of the business cycle—this study analyzes how households adjust their grocery shopping behavior. The authors observe more than 5,000 households over eight years and analyze shopping outcomes in terms of what, where, and how much they shop and spend. Results show that micro and macro conditions substantially influence shopping outcomes, but in very different ways. Microeconomic changes lead households to adjust primarily their overall purchase volume—that is, after losing income, households buy fewer products and spend less in total. In contrast, macroeconomic changes cause pronounced structural shifts in households' shopping basket allocation and spending behavior. Specifically, during contractions, households shift purchases toward private labels while also buying and consequently spending more than during expansions. During expansions, however, households increasingly purchase national brands but keep their total spending constant. The authors discuss psychological and sociological mechanisms that can explain the differential effects of micro and macro conditions on shopping behavior and develop important diagnostic and normative implications for retailers and manufacturers.

A randomised experiment of health, cost and social norm message frames to encourage acceptance of swaps in a simulation online supermarket

Offering lower-energy food swaps to customers of online supermarkets could help to decrease energy (kcal) purchased and consumed. However, acceptance rates of such food swaps tend to be low. This study aimed to see whether framing lower-energy food swaps in terms of cost savings or social norms could improve likelihood of acceptance relative to framing swaps in terms of health benefits. Participants (n = 900) were asked to shop from a 12-item shopping list in a simulation online supermarket. When a target high-energy food was identified in the shopping basket at check-out, one or two lower-energy foods would be suggested as an alternative (a “swap”). Participants were randomised to only see messages emphasising health benefits (fewer calories), cost benefits (lower price) or social norms (others preferred this product). Data were analysed for 713 participants after exclusions. Participants were offered a mean of 3.17 swaps (SD = 1.50), and 12.91% of swaps were accepted (health = 14.31%, cost = 11.49%, social norms = 13.18%). Swap acceptance was not influenced by the specific swap frame used (all p > .170). Age was significantly and positively associated with swap acceptance (b = 0.02, SE = 0.00, p < .001), but was also associated with smaller decreases in energy change (b = 0.46, SE = .19, p = .014). Overall, offering swaps reduced both energy (kcal) per product (b = -9.69, SE = 4.07, p = .017) and energy (kcal) per shopping basket (t712 = 11.09, p < .001) from pre- to post-intervention. Offering lower-energy food swaps could be a successful strategy for reducing energy purchased by customers of online supermarkets. Future research should explore alternative solutions for increasing acceptance rates of such swaps.

Clustering and Profiling Consumer Buying Behavior

This paper aims to identify the factors that affect consumers' buying behavior towards goods of consumers' shopping basket to classify them into groups according to their similar buying behavior patterns and to profile each group of consumers. A primary survey conducted to 242 consumers in Greece. Principal component analysis (PCA) conducted to identify the main factors that affect consumers purchasing behavior. Cluster analysis performed to classify consumers into groups with similar purchasing behavior whilst discriminant analysis conducted to check cluster predictability. Nonparametric tests are performed to profile each group of consumers according to their demographic characteristics and other factors. PCA identified six main factors: (1) price, (2) entertainment during shopping, (3) advertisement, (4) public relationships, (5) product features, (6) promotion activities. Cluster analysis classified consumers into three groups: (1) advertisement-orientated consumers, (2) promotion-orientated consumers, and (3) entertainment-orientated consumers.

Effects of Labelling and Increasing the Proportion of Lower-Energy Density Products on Online Food Shopping: A Randomised Control Trial in High- and Low-Socioeconomic Position Participants

Reducing the energy density (ED) of product selections made during online supermarket food shopping has potential to decrease energy intake. Yet it is unclear which types of intervention are likely to be most effective and equitable. We recruited 899 UK adults of lower and higher socioeconomic position (SEP) who completed a shopping task in an online experimental supermarket. Participants were randomised in a 2 × 2 between-subjects design to test the effects of two interventions on the ED of shopping basket selections: labelling lower-ED products as healthier choices and increasing the relative availability of lower-ED products within a range (referred to as proportion). Labelling of lower-ED products resulted in a small but significant decrease (−4.2 kcal/100 g, 95% CIs −7.8 to −0.6) in the ED of the shopping basket. Increasing the proportion of lower-ED products significantly decreased the ED of the shopping basket (−17 kcal/100 g, 95% CIs −21 to −14). There was no evidence that the effect of either intervention was moderated by SEP. Thus, both types of intervention decreased the ED of foods selected in an online experimental supermarket. There was no evidence that the effectiveness of either intervention differed in people of lower vs. higher SEP.