The Hidden Geometry of Retail

Retail store managers and practitioners have long operated with a simple principle: place related products close together and customers will buy more. It is intuitive, practical and almost universally practiced. But it is also incomplete. For decades, retailers stressed over product placements within departments. Multiple research articles have been published that pushed for “eye level is buy level”, endcap promotions, and using planograms for shelf rearrangements but little empirical work exists in the space exploring a fundamental question: how far apart should entire departments that contain these products be with respect to each other.

Take an example of bread. Most stores have a counter baking fresh bread. But did you also notice that bread exists as a separate department in an aisle far apart placed with other products? How different is “corporate” bread from a “store baked” bread. In a normal world they are substitutes, but are they in a supermarket? Think again.

My new research “The Impact of Inter-Departmental Distance on Joint Sales in Retail Stores”, written with Praveen K Kopalle (Dartmouth College, USA), Stephanie M Noble (University of Tennessee), Jens Nordfält (University of Bath) Dhruv Grewal (Babson College, USA) and forthcoming in the Journal of Marketing, reveals that distance between departments and the combined sales of the two departments follows an inverted-U pattern. That is sales peak between departments when they are not closest together but moderately distanced. 

When departments sit next to each other, shoppers tend to view them as substitutes. The bakery next to the desserts section? Consumers unconsciously frame them as competing options. “I’ll grab something from one of these” becomes the default option and combined purchases decline. When departments are far apart, a different problem emerges. Consumers view them as unrelated. Even naturally complementary categories fail to trigger joint purchases. Only in moderate distances, the related but distinct aspect stands out driving joint purchases. This is where the products appear similar enough to trigger associative thinking but different enough to warrant separate purchases.

While we understand that reconfiguring a store requires significant investment and loss of revenue due to renovations, two simple factors can be considered.

1. Non-identical layouts: When two departments have different layouts, say one is grid layout and the other free flow, the inverted U effect is strengthened, that is, visual distinctiveness reinforces related but different aspect.
2. Different in the number of products per department: When one department offers significantly more variety than another, the distance effect intensifies. The contrast makes each department’s unique value more noticeable.

Our findings emerge from a multi-method research approach, combining store-level field data, controlled lab experiments, and optimization. We examined 64 stores that consist of 32 supermarkets and 32 hypermarkets from a retailer in Sweden for 52 weeks. Using Baron solver, which applied advanced global optimization techniques, we calculated the optimal store layout that maximize total revenue. 

The potential to increase revenue for supermarkets is almost as high as 9.5%. For example, for a supermarket generating €500,000 in weekly sales, the average of 4% improvement translates to €20,000 per week which is over €1 million annually. This is important because these gains require no marketing spending or price reductions which is typically done by stores but purely emerging from rearranging existing departments. We converted our code into action and created a store layout by collaborating with an architect which can be used by retailers when planning their renovations. 

For retailers operating in fiercely competitive environments where margins are thin, layout optimization is an understudied and unexplored lever. The fix is not complicated. Audit existing layouts, track combined sales of departments rather than sales of each department, and invest in optimization tools. The geometry of a store is far more than a logistical matter; it is a strategic asset hiding in plain sight.