Compared to traditional Real Shelf or Online Shelf tests, Virtual Reality (VR) shelf tests are a boon for researchers who need to test the effectiveness of many shelf layouts in a standardized environment or need to test the same shelf in multiple regions. However, how reflective of real behavior are they?
To answer this question, we conducted research-on-research to compare a real shelf with Virtual Reality and online shelves. We used a simple 3-cell monadic design in which a total of 225 participants shopped each shelf in the Explorer Shopper Lab while wearing Tobii eye tracking glasses.
The data shows that because the life-sized VR environment allows shoppers to interact with products in a manner similar to real life, e.g., bending down, walking back and forth, and focusing on landmark brands, Virtual Reality testing accurately replicates real shopping. The two methodologies generated very similar shop times, units purchased, viewing locations, and more. (See results in the table shown below.)
However, online shelf results differed in important ways from the other approaches. Specifically, the online shelf generated a longer average shop time. Online participants also selected a greater number of brands positioned at the center of the shelf and spent more time viewing the bottom of the shelf.
Virtual Reality Shelf
Average shopping length
Average units purchased
Percent of units selected from bottom shelf (of four shelves)
Stopping power of Brand B
(Visibility indexed to 100)
Closing power of Brand B
(Sales Performance indexed to 100)
In addition, Brand B, which was located on the second shelf, elicited more eye gazes and purchases than both Real Shelf and VR. This can be seen in the Eye-Tracking heat map below in which increased gazes are present in the center and bottom shelves.
It is clear that virtual reality accurately replicates real shopping behavior. Unlike Online Shelf tests, VR shelves require the same product interactions – bending down, walking back and forth, and focusing on landmark brands. The results also reveal that Online Shelf studies have potential shortcomings, most notably that they may inflate purchase rates, especially of brands located in the middle and lower areas of the shelf.
Where highly accurate shelf test results are necessary, Virtual Reality offers researchers improved external validity over Online Shelf tests.
Will Cornish is a Vice-President at Explorer Research and works with a variety of CPG, Retailer, and other clients. He also heads up Explorer’s Media Testing Practice. He has 12 years of quantitative and qualitative research experience and before joining Explorer he was at TNS Canada.
Explorer Research applies behavioral science to research methods, with an emphasis on in-situation testing to help clients uncover insights related to packaging, advertising, media, customer experience, and user experience. Using an array of conscious and nonconscious research measures, including eye-tracking, facial coding, EEGs, GSR, VR, and more, Explorer Research helps their clients understand, predict and influence consumer behavior. Follow us on Twitter or LinkedIn.