CPGmatters: Nestlé Deploys Virtual Shelf Tests to Revitalize Ice Cream Category
Through a special arrangement, presented here for discussion
is a summary of a current article from the monthly e-zine, CPGmatters.
Sometimes
the biggest ideas come in small packages. A few years ago, Nestlé decided
to jumpstart the slow-growth U.S. ice cream category. So it turned to ice cream
cups. The consumer benefits were obvious: convenience/portability, portion
control, and the ability to buy a variety of flavors to please all family members.
Soon
after Nestlé expanded its offering in early 2008, sales of single-serve
cups grew briskly. Last summer, Nestlé was making plans to introduce
fifteen SKUs of ice cream cups across their Edy’s/Dreyer’s, Haagen-Dazs
and Skinny Cow brands for 2010. The marketer needed to address two key questions:
- How should the cups be deal priced – $10 for 10 or 99¢ each?
- Should the ice cream cups be merchandised with their parent full-size offerings
to create a brand-blocked family or together in dedicated doors as a cup
destination?
"We couldn’t get into a store for testing because we didn’t
have the offering or the time," said Russ Onish, director of category leadership & shopper
insights at Dreyer’s Grand Ice Cream, at the recent Shopper Insights in
Action conference in Chicago. "So we set up a virtual store environment
to let shoppers decide. Let them shop a full 20-door aisle of ice cream set different
ways, measured their response in purchasing online, and then made our decision."
Kansas
City-based Decision Insight and Nestlé conducted the study in
multiple virtual supermarkets and regions: Safeway in Southern California and
Kroger in Michigan and Ohio. To create awareness and simulate promotion, single-serve
cups were initially advertised in the store circular.
Here are the results
of the virtual testing:
- More revenue results from grouping all of the cups behind dedicated doors
(but Skinny Cow cups fared better when it remained with its brand family).
- Dedicated doors (sans Skinny Cow) drive multiple cup purchasing and incremental
variety seeking.
- Shoppers buy more cups in a transaction when priced at 10 for $10 than
when the price is 99 cents per cup.
Within six months of completing the virtual shopping study, Nestlé implemented
the recommended solution in nearly every one of its retail customers. Ninety
percent of stores now stock cups together. Sales of all cups increased 53 percent
for the 12-weeks ending June 19, 2010 (excluding Walmart and HEB), according
to statistics from The Nielsen Company.
Eighty-five percent followed the recommendation
to keep Skinny Cow as a separate brand block, as a result Skinny Cow sales
have more than doubled in those stores.
Discussion Questions: What are the advantages and disadvantages of virtual
shopping testing of a product versus in-store testing? What steps should
brands take to make up for any shortcomings from virtual testing?
In the interest of full disclosure, VR is how I make my living. VR testing is faster and cheaper than an in-store test, and can often permit the researcher to test more options in more types of environments than an in-store test would permit (due to cost and timing reasons). It is, however, a simulation, and that always needs to be kept in mind.
We have lots of validation data and experience has taught us what will and won’t work in VR and how to either counteract a problem or avoid the research project (we hate to reject a project, but we will if we’re convinced it can’t be done).
When we do a study that may be influenced by the VR methodology, we encourage our clients to test the results live with a retailer or two. In the past few years, we’ve seen a number of grocery, drug, and mass retailers willing (even eager) to try something new. This provides a sanity check for overcoming issues like “it’s not real money they are spending,” which could have been an issue with the Nestle study.
Virtual store testing is a cost effective way for a brand to do initial testing, especially when the item is not yet available. However, the results will not always be as obvious as they were for Nestle and their Ice Cream shelf study. Environment has a major impact on how a consumer shops and you can’t replicate that virtually. Does the shopper have an infant or small child with them, are the aisles busy, what type of music is the store playing, does the time of day effect purchase behavior, and is the lighting good?
Once you have your initial read from the virtual test, a store test should follow to validate your findings before rolling it out to all stores.
We have been involved in a number of virtual shelf tests for various categories and products. They are much faster and easier than store tests. They tend to work better in frequently purchased categories where the shopper can also draw on store experience. They work less well in situations that are need state or environment driven. For instance, it is difficult to simulate hunger/thirst or the need to solve a problem like an odor in the home.
Overall, we have found that they can be quite useful in gaining an initial read on many new concepts. However, they typically need to be corroborated by in-store work.
Steve, of course, knows his business and Decision Insight is a reputable company, so Nestlé was in good hands. However, the suggestion of validating the virtual result by an in-store test was already attempted by Nestlé and rejected by retailers. That’s why they went virtual. Even with positive results, the retailers will be very skeptical about changing their shelves as they don’t trust the manufacturers. Plus, what effect did the advertising have on the jump in sales and can that be duplicated nationally?