Category management is becoming both more complex and more critical to profitability. During a recent webinar hosted by the Category Management Association (CMA), representatives from Walmart and Procter & Gamble (P&G) outlined their own efforts at improving their assortments, reducing SKU counts to gain share, and how technology and training have helped reduce costs and improve margins.
The webinar, "Discover the State-of-the-Art in Assortment Customization," was held in February. While the companies involved targeted their conversation at the consumer packaged goods industry, there are some key lessons (and warnings) that are relevant for aftermarket suppliers and retailers as well.
According to The Theory and Practice of Optimizing Assortment and Space, a white paper on the topic written by CMA, manufacturers increasingly have to tailor their assortment on a store-by-store, retailer-by-retailer, and SKU-by-SKU basis, a complex endeavor that can quickly overwhelm available resources for both suppliers and retailers.
"If you are in 20 different geographic markets, but also have three to four clusters of stores, that means your external market has those markets times the clusters, giving you potentially hundreds of external markets to deal with," says Gordon Wade, director of best practices at the Category Management Association.
The keys to succeeding at ensuring you have the right assortment at each location will be having trained staff with sophisticated skills in category management; a structured process for ordering/ranking the SKUs and markets; new software capabilities that can create thousands of individualized, store-specific versions for assortments and planograms; and robust data and analytics tools that can help identify and classify "micro markets" unique to multiple retailers.
Preparation is critical, according to Michael McGuire, senior director of modular development at Walmart. Manufacturers and retailers have to know their market and their competition in that market, and understand regional dynamics and regional competitors. Companies also have to know the product, both the science/engineering behind it and how the customer views and experiences the product.
"And get out into the market and into the stores," McGuire says. "I'm a huge advocate for data, but there is no replacement for talking to associates, and getting feedback from customers."
It's also important to know how your customers shop a given category. "You really want to understand how they make their decision on where to shop, and how they select an item," McGuire says.
Fewer SKUs, more sales
The data from both Walmart and P&G indicate that reducing SKUs in a given category can help improve sales for the retailer and increase share for the supplier.
To make this SKU reduction work, you have to determine each SKU’s "transferable demand," or the likelihood that the customer will buy a different SKU in the same category, rather than go to a different retailer to obtain the eliminated SKU item.
For highly fragmented categories, CMA recommends a demand-based clustering response with unique assortments and multiple planogram variations based on specific demand drivers at the store or store cluster level.
Store clustering also helps identify the competitive or external market for a cluster. That way, manufacturers can identify SKUs that are doing well in the demand cluster, but are absent from the specific retailer's assortment. At that point, companies face the issue of transferable demand. Do those items not in the retailer's assortment represent demand unique to those items, that can't be transferred to a comparable item in the current assortment?
Making those determinations is only possible using advanced category management software tools. There are algorithms available that can calculate transferable demand across a spectrum of SKUs to identify items not in the assortment that deserve inclusion. Kantar, APT, 5th Dimension, JDA, and other vendors have such solutions
In the case of P&G, the company used JDA's Planogram Generator to produce thousands of store-level planograms at Walmart in just a few days. According to CMA, the ability to host or access tools and data in the cloud has been a boon to these data analytics efforts, because outputs can be easily shared and software improvements are more cost effectively accommodated.
According to Tom McDonald, team leader for P&G's Walmart and Sam's business, software alone was only part of the solution. The company had to prepare internally to create customized planograms efficiently, and that meant having the right data on the stores and movement by store, as well as developing store clusters based on multiple factors.
"Our next step was to develop gold standard planograms suggesting basic assortments for the key retail clusters that we identified," McDonald says. "Then we ranked the SKUs by segment based on Walmart’s POS. As a final step we created the external markets for each cluster and identified the popular SKU’s that were not in the gold standard or in Walmart’s current assortment."
In many cases, P&G found that it needed to reduce its assortment. McDonald described a test in which shoppers were presented with two different tables filled with different jams. One had six jams, while the other had 24 jams. More people stopped at the 24-item table and lingered longer, but more shoppers purchased at the six-jam table. "Extensive choice can cause shoppers to delay or forgo purchases," McDonald says. "Adding items is not a recipe for growing a category. There is a law of diminishing returns."
In fact, P&G has found that retailers that expand assortments often find themselves losing share and money. "We knew when we started this that we had to thin our assortment to make it easier for shoppers to find the products they want," McDonald says.
At P&G, the company has found that downsizing SKU counts has driven sales and reduced inventory costs, but only where the right data and expertise were applied to choosing which SKUs to eliminate. "We've tested upwards of getting out of 30 percent of SKUs in a key category, and reduced inventory by 5 to 8 percent at the retailer level, and have grown those test stores faster than the control stores, even though we have fewer SKUs," McDonald says. "You have better in stocks, and less shopper confusion."
While software played a role at P&G, McGuire cautions that you have to find the right software for your operation.
"Cool or interesting is not the same thing as effective," McGuire says. "You want to make sure any software you are using is practical for achieving your goals for the category. One size almost never fits all, and each category has a different creative technology solution. We're most successful when we pair a deep knowledge of the customer and the category with the technology."
All the technology in the world doesn't help if your staff isn't properly trained in assortment and category management, and how to use the data. "You have to get your people to a skill level at which they can help the retailers further drive share," McDonald says.
Subscribe to Aftermarket Business World and receive articles like this every month….absolutely free. Click here.
