Assaying the impact of in-store product sampling

In "An Assessment of When, Where and Under What Conditions In-Store Sampling is Most Effective," the three authors -- Sandeep R. Chandakula of Singapore Management University, Jeffrey P. Dotson of Brigham Young University, and Qing Liu of University of Wisconsin-Madison -- find that sampling has both an immediate, if short-term, effect and a sustained impact on sales, but that the impact varies according to the size of the store conducting the event. They also found that repeated sampling for a single product produces increased returns and that sampling tends to expand a category rather than purely substitute for another product.

The authors created their model from a series of six scanner datasets collected on four different snack product categories, including both new and existing products, over a period of weeks. They performed a series of regressions testing a number of variables, such as store size, price, and ACV, or all-commodity volume, an indicator that helped them determine that smaller stores with a smaller assortment of products seem to benefit more from sampling than larger stores.

Of key importance is the finding that in-store sampling outdoes simple end-of-aisle displays, the latter of which lose their oomph after two weeks whereas sampling's effects, in addition to beating displays for immediate sales results, linger for many weeks. "Taken collectively, we can infer that the total effect of sampling, as measured by an incremental lift in sales, is much larger than that of in-store displays," the authors write.

Manufacturers will also appreciate the valuable rule of thumb revealed by the data: if the incremental cost of a sampling event exceeds 15 times the unit price of the product, the event ceases to be profitable. Other key findings: execution of a sampling event by one brand led to an increase in sales for all brands in the category, which they suggest should interest the category manager seeking to find ways to increase total category profitability, not the individual performance of a single item or brand.