Interesting article on how automated audience measurement systems are being used in a fully monitored research environment to help organizations understand the effectiveness of their products in retail stores. You can read the full article here.

Frito-Lay may dominate the snack aisle of the grocery store, but it is always looking for a competitive edge to increase sales.
What’s providing that edge nowadays is its S.M.A.R.T. Learning Center in Dallas, a 15,000 sq. ft. simulated supermarket that tests the effectiveness of various marketing solutions to drive impulse purchases. Through automated observation and other research methods, the marketer studies the behavior of actual consumers as they shop the aisles and react to different merchandising tactics.
Frito-Lay built the learning lab to answer three fundamental questions:
- What would lure shoppers to the snack aisle?
- What merchandising solutions would encourage people to go down the snack aisle?
- Once there, how could the shopping experience be improved?
Clancy was in charge of a series of tests in the learning lab last year that laid the groundwork for the ongoing research that continues to pay off for Frito-Lay.
“The key thing we struggled with in the past was how shoppers would react when they were actually in that environment. You can observe them in stores, but this was the first time we could do it on our own. We could create whatever we wanted to. We could change the merchandising, the product, whatever we wanted.
“My job was to think of and implement new research methodologies,” he went on to say. “We wanted to understand what happened to shopping behavior and the perception and purchase of our brands when we made certain changes to our aisle”
Frito-Lay sought to determine the effectiveness of displays in attracting shoppers and to understand the relative performance of different merchandising concepts in driving impulse purchases. Overall, the simulated store served as a platform for transforming
in-store marketing.
Clancy said the in-store behavior of shoppers reflects their actual response to different elements of marketing and merchandising. So, an important element of the lab was measuring shopper responses using various solutions such as transaction tracking (loyalty data, sales, household panels), survey research, cart tracking, manual observation, and automated observation.
“I wanted to use as many multiple methodologies as possible. We didn’t know which ones would work or how they would work. So we had quantitative, qualitative, and observational. We used five methodologies to see which one would work better for our needs. But the overall learning was that putting them all together worked best. Our learning was so much deeper that it would have been had we just gone down one path.”
Clancy singled out the important contributions of VideoMining, State College, Pa. The firm’s in-store technologies and analytics form a scalable measurement platform designed to deliver precise metrics on shopper response at every retail touchpoint.
“VideoMining could quantify all of the shopping and tell me that people were lingering a lot more at certain points in the aisle. ‘Heat maps’ showed where shoppers were spending more time. It really helped us understand which merchandising was working and which was not.
“The way we merchandised the aisle made shoppers look for new things and spend some time in different sections – really browse as opposed to search,” Clancy said. “Browse is a good thing, and search is not a good thing. That was a key learning for us. Knowing where shoppers were going in the aisle helped us optimize the product set and sometimes change the flow a little bit, too.”
He listed the following benefits of studying shoppers in the simulated supermarket:
- Understanding the “shopability” of various marketing strategies
- Deeper understanding of shopper behavior
- Understanding display effectiveness by type and location.
- Appreciating the power of observational techniques.