Washington D.C. – The New York Times is reporting that many retailers are going back to the true and tried strategy of stocking their shelves, aisles and backrooms with more inventory. As we discussed before, the notion of shoppers preferring to have less choice is a myth. Many retailers learned this the hard way in the last 3-5 years: SKU rationalization efforts have not worked as well as merchants at the leading retailers have hoped.

The Paradox

Many retailers are bringing back pallets, end caps and checkout lane displays to entice shoppers to spend more. It is a surprising turn of events for several reasons. The chief reason is that many of the same companies were touting their “clean look” and “uncluttered feel” strategies as recently as twelve months ago. This is especially surprising because many of these same retailers surely tried their “clean” and “uncluttered” looks before rolling them out nationwide. Many even ran in-market tests to ascertain their strategies. Why then did the results of those tests not translate to national rollouts?

On Accuracy

Tests need to be designed properly to provide accurate results and inform go-forward decisions. Any test needs to have the right number of stores or markets to provide meaningful recommendations. If a test is too small, then it will be challenging to confidently measure the impact. 5-store or 2-market tests are not tests at all in the sense that they cannot be relied upon to guide rollout decisions. These are more like operational trials that can inform the operational viability of the idea, but not provide enough statistical significance to bank company’s future on.

Secondly, test stores and markets need to be selected to be representative of the network as a whole across key dimensions. If the test is biased in some way (e.g. we only test the “clean look” concept in downtown Dallas), the results won’t hold when we extend the idea across the network.

Lastly, measuring the impact of any test requires rigor and analytic sophistication.

• Proper control groups have to be used to account for seasonality and economic fluctuations. It is not enough to show that performance has improved on a pre vs. post basis (or this year vs. last year). Only using a well-designed control group can tell us what the true, incremental impact of a program actually is.
• Understanding the true impact of a test requires additional measurements beyond the specific product and metric tested. For example, when testing the new store format it is not enough to look at sales and customer satisfaction. One must examine impact on unit sales, average basket size, customer count, gross profit, etc. Furthermore, one needs to understand how these changes happen at the department or category level, not only at the total store level, and segment results by customer types.
• Finally, any test measurement must properly identify and suppress outlier observations. Results that are not representative of the rollout should not be used in making the rollout “go” vs. “no go” decisions.

It is not enough “to test” and “to measure,” but you actually have to have the expertise and the tools to be able to do this accurately. If you don’t, you should be working with APT to take your testing capability to a different level. Stop relying on your gut-feel. Instead embrace the power of data-driven decision-making.

Learn more

Learn how APT’s Test & Learn^TM helps grocery and c-store retailers make decisions on capital allocation strategy with great confidence.