Solving the Shelf Puzzle in Retail as an FMCG Manufacturer Under Limited Shelf Capacity
I conducted a retail shelf space optimization project in one of the biggest snacks manufacturer in Turkey for a key retail account, and here is everything I learned about shelf space allocation.
The competitive landscape for consumer packaged goods is changing rapidly. Shoppers are now more digitally empowered and have adopted cross-channel shopping behavior. This means that they are shopping online, in stores, and through other channels. As a result, retailers are competing for shoppers’ every single penny in every neighborhood.
Especially in a country like Turkey where consumer habits are gravitating towards more affordable alternatives, packaged snacks are accelerating in demand. In such environment, manufacturers are constantly battling for every centimeter square in retail shelves.
From a game-theoretic approach where every player aims to maximize profit, it is crucial for both manufacturers and retailers to ensure a fair allocation in shelves. That was the starting point for my project, checking whether the products on the shelves are displayed fairly.
In this competitive environment, one of the prominent strategies of these snacks companies, to make annual agreements with retailer to ensure a certain amount of shelf space. This, on the other hand, was one of our biggest obstacles in this project since we wanted to ensure a fair arrangement between brands according to their market shares.
We began our with situational analyses with determining the capacities of the shelves and checking the current displayments of the products. The main takeaways from these analyses were:
- Bigger manufacturers were getting less space share than their market shares, whereas it was the opposite for smaller brands.
- Minor brands who have less market share than 1/shelf capactiy, were listed in the retailer, hence have a shelf space.
One of the main reasons of the situation of “1.” was “2.”, which also led us to a suboptimal result. Due to competition regulations, we need to accept all of the listed products as listed, and could not offer any deslisting from shelves.
Under these assumptions, I created an Integer Programming model to allocate frontfaces (the instance of an SKU is facing front next to each other in a shelf) fairly with their respective market shares. An important constraint for this model was to ensure at least 1 frontface to each SKU due to delisting not being an option.
I needed to ensure that we maximize the utilization of the shelf in two ways:
- area used/total shelf area: For this objective, I needed to know the exact measurements of the shelves and every SKU listed, which we did not. To overcome this obstacle, I benchmarked the current planograms as 100% utilized, and constructed linear equations by dividing SKUs in 3 categories by their weights, a parameter I could obtain: small, medium, large.
- market share of an sku/brand in that shelf/space share of that SKU/brand in the shelf: This is the main objective to ensure a fair shelf space construction.
With respect to these objectives, we determined frontface numbers for every SKU in every planogram.
The second part of the project was to place these “frontfaces” into the shelves. This is rather the psychological part where you cannot measure the performance of. From my literature research, I learned that there are several factors that influence consumer behavior and purchasing decisions at the shelf. These include product visibility, brand loyalty, packaging design, price, and promotions. To optimize shelf space allocation, it’s crucial to consider these factors and how they interact with each other.
One approach to optimize shelf space is to use category management principles. This involves grouping similar products together and ensuring that they are displayed in a logical and intuitive way. For example, in a chcolates category, you might group tablets together and bars together. Within each category, you can further optimize shelf space by arranging products based on their popularity and profitability.
Another approach is to use data analytics to understand consumer behavior and preferences. By analyzing sales data, you can identify which products are selling well and which ones are not. You can also analyze consumer demographics and preferences to understand which products are most appealing to different segments of the population. This information can be used to optimize shelf space allocation by placing popular products in more prominent locations and less popular products in less prominent locations.
In our project, we used a combination of these approaches to optimize shelf space allocation for our client’s products. We grouped similar products together and arranged them based on their popularity and profitability. We also analyzed sales data and consumer preferences to identify which products should be placed in more prominent locations. Further, important constraints to consider were placing children’s products into the bottom shelves and premium products into the top shelves.
The classification of SKUs by their weights was also valid in this part. In that way, I could ensure when implemented, the products will fit in the shelves and the empty spaces were minimized.
Overall, our project resulted in a significant improvement in shelf space allocation for our client’s products. By using a combination of category management principles, data analytics, and integer programming, we were able to optimize shelf space allocation based on market share while still ensuring fair treatment for minor brands. In the second part, I will be talking about challenges in the implementation and performance analysis of the system.
