One of the most persistent trends of the last 30 years of CPG marketing has been SKU (Stock-Keeping Unit) proliferation – the addition of more products and a variety of products based on changes to the market, such as adding several new flavors of a top-selling beverage. However, this trend is coming to an end. 

Over the next five years, we will see a contraction in the number of SKUs in the typical food retail store, driven by several factors: retailers’ desire to free space for more service-intensive, higher-margin products; the need to use space more efficiently with faster turning, more popular items; and to reduce supply chain and store labor costs.

The reduction will be concentrated in the 20 or so large categories containing the most SKUs with duplicate attributes, such as those stocking 60 SKUs of vanilla ice cream or 30 SKUs of blueberry yogurt. A shopper seeking vanilla ice cream or blueberry yogurt simply does not need this many options.

As the person who literally wrote the initial efficient item assortment protocol for the CPG industry as part of the development of category management in 1994, I can assure you that retailers can provide adequate attribute choices in a multi-segmented category to meet the needs of virtually all shoppers.

Too many SKUs

The challenges caused by SKU proliferation are quite real, but not well understood. The biggest problem is this: you cannot give a SKU less than one facing on the shelf. This facing represents a certain proportion of the total facings available in the category, but many small-volume items do not earn that percentage of space. In fact, in many categories, these multiple smaller volume SKUs actually crowd out popular fast-moving SKUs and cause items to be out of stock, particularly in some historically space-constrained categories such as the frozen and refrigerated categories.

The supply chain problems caused by the pandemic highlighted the problem of SKU proliferation and the resulting out of stocks, which has added to retailer’s interest in SKU reduction. Their concerns are well founded, and evidence exists supporting the value of a targeted reduction in the number of SKUs in certain categories.

A case study in SKU reduction

In 1994, in conjunction with the creation of the efficient item assortment protocol, I and some former colleagues at The Partnering Group conducted two tests of aggressive SKU reduction. The first was done in the carbonated beverage category at Dominick’s in Chicago. The store carried 400 SKUs in the category. We removed 20% of the SKUs, which equated to 80 items. Interestingly, the result was that volume in the category actually INCREASED by 3%. We interviewed shoppers to understand what was causing this phenomenon.

The responses we received were quite logical. Shoppers actually thought that the category contained more SKUs because they were able to find their preferred item easier in the reduced assortment. We reasoned that the small slow moving SKUs were crowding out the more popular SKUs, primarily because retail aisle clerks were facing out of stocks of those more preferred SKUs than with those that were less preferred. Shoppers confirmed they could more easily find the SKUs they preferred.

Shortly after that test, we conducted a similar test in the ice cream category of a Stop & Shop where they also stocked 400 individual SKUs and again removed 20% of the SKUs much to the consternation of the category manager. He was terrified that he would be deluged with requests for the removed items by outraged shoppers. In fact, he only received one complaint and that was caused by a short shipment of an item that we had kept in the assortment but was out of stock because of a supply chain problem.

While the tests we conducted required expensive and time-consuming resets, today any retailer or category captain wishing to quantify the effect of SKU reduction could avail themselves of digital twin research approaches available from providers such as InContext Solutions.

To select which items should be delisted for these tests, we recommended retail category managers utilize the household panel data which identifies critical metrics. By focusing on these metrics, we were able to eliminate items with low rates of loyalty, very low rates of exclusive use, high rates of substitutability, and purchase by shoppers with low category worth.

Over the years, analogues of these attribute metrics have been included in analytics by many major solution providers such as IRI, NielsenIQ, and Kantar’s rRch Mix solution. The details for creating these critical metrics were included in the efficient assortment protocol that I wrote and subsequently have been included in several white papers over the years.

What retailers can do

Here’s what you could do to quantify the risks of delisting your items or the value to the category by removing or adding certain items to the assortment.

  • Use household panel data to develop the metrics of item loyalty, exclusivity, substitutability, and shopper worth for your items. Compare your items to other items in the category.
  • Conduct digital twin category research to quantify the risks and rewards of reducing specific SKUs from a category assortment.
  • When justifying the inclusion of a new item in the assortment, establish that the new item is truly unique in its attributes such that it appeals strongly to a shopper’s desire unavailable in the current assortment.

Following these guidelines will help you to develop categories that are not only easier to shop and manage, but are also much more profitable. Go big, by thinking small!

Hi there 👋
Want the inside scoop on all things CPG?

Get the latest CPG and retail insights, trends, and business best practices sent directly to your inbox, every week.