What’s Next for Category Management Analytics?

In the mid-1990s, with the availability of a plethora of syndicated data collected each time a product scanned at a register, an analytic approach to promotion, pricing, and product assortment emerged; it was called “Category Management.”

The syndicated data provided detailed information about the products consumers purchased in-store including pricing, (TPR, promoted, non-promoted), displays, ads, promotional lift, etc. This data helped build a history of promotions to optimize the promotion strategy, such as the depth and frequency of promotions. The data could also feed into pricing strategy optimization (hi-lo, EDLP).

The data were also quite effective in conducting SKU rationalization for determining the optimal product assortment and plan-o-grams based on movement, a number of facings, days of supply, and turnover rates. Analysis of this data also played a significant role in item forecasting. The more accurate my forecasts were when working with a fresh product, the less chance I had of over/under-manufacturing a product.

 In 1996, I was promoted to the newly created position of “Manager of Category Development” with The Minute Maid Company, a division of The Coca-Cola Company. I participated in the second category management training class conducted by The Coca-Cola Company.

Analyzing the data was tricky as not all accounts provided syndicated data, and today’s artificial intelligence Catman programs did not exist. 

We “pulled” raw data from the IRI and Nielsen databases and dropped it into Excel and PowerPoint workbooks and presentations. It was laborious, and preparing for an account’s category management review took upwards of six weeks to capture and format the data for analyzing and making recommendations.

Related Article: 5 Pillars of Organizational Culture in Hispanic Retailers

I’m forever grateful for the time I held this position. Being there at the beginning of the category management movement was pivotal in my career. It allowed me to experience a move into fact-based selling, which did not fully exist before the availability of scanner/syndicated data.

In category management, syndicated data captures what consumers are purchasing in the store and taking home to consume presumably. The only data available before scanner data was warehouse withdrawal reports, called “SAMI” reports, “Selling Areas Marketing, Inc.”

Warehouse withdrawal information can be misleading as it captures products leaving a warehouse for any reason such as: diverting, store forward buying/loading, displays, etc. Eventually, that product might make its way into a consumer’s home, but the ability to analyze why and how it got there is lacking with only the use of warehouse withdrawal information. It’s impossible to know pricing and the impact it might have played; or how in-store merchandising such as displays and ads impacted a consumer’s decision. It helps determine a promotional and pricing strategy.

The last couple of years has turned the world of category management/forecasting upside-down

The role historical analysis play is such a large part of category management, but due to the consumer behavior changes, as a result of the pandemic, how might a manufacturer forecast for the future as the world starts to return to somewhat of a pre-pandemic normal?

How can manufacturers forecast what consumer behavior changes will be long-term? Will at-home eating habits change as more employees leave their home offices and return to office buildings and students return to school?

There is a problem with syndicated data, and that is it does not capture purchases made online. With online shopping increasing during the pandemic, how to fill the data gap to establish an optimal promotional strategy in 2022 and beyond?

Consumer panel data reported on e-commerce purchases will be a trend in 2022. A few vendors provide broad consumer panels to capture this void in syndicated data. A new set of factors will emerge that manufacturers will analyze to determine promotional efficiency.

As online shopping increases, those factors will move from measuring the promotional impact (lift) of ads, displays, and TPRs to new factors.  With many retailers moving away from ads and with no syndicated data available for online shopping, what might the new promotional efficiency factors be based on? How will forecasters account for this lack of data in their forecasting analysis?

We’ve had a rapid increase in online shopping and e-commerce platforms and providers in the last couple of years. What is next for category management analytics? I believe that in 2022, we will see an influx of innovative and accurate ways to capture online sales and the factors which drive those purchases.