Why In-Store Research?

Recently I was asked why I have chosen to conduct or propose so many projects under the in-store protocol. I can give a lot of reasons, but probably the most important is that the in-store research environment frequently gives different results than the same study conducted in the conventional research environments such as Mall or CLT testing. When I have gotten different results, the client always accepts the in-store results over those obtained through other sources. Also I have data that shows when there is a conflict between the in-store data and data collected through typical testing facilities, the in-store data reflected more accurately what happened in the market introduction.

I have never really had the opportunity to design a test directly comparing the validity of conflicting results since I was never able to convince a client to pursue an initiative that the in-store research said should not be pursued. [This is similar to the problem I have with companies that say they have validated a particular protocol but never seem to have their clients put initiatives that failed in their research into the market to determine if they are actual failures (the Beta Risk).] I have, however, taken failures in conventional market forecasting methods and conducted an in-store study such as a Disposable Test Market and found the brand to be a success and it was ultimately confirmed in the market. I have also researched a brand that conventional market forecasting research predicted a winner and failed in the market. The in-store research not only predicted failure but identified the basis of the failure.

The following paragraph was taken from my Views of August 25, 1996. This is an excellent example of the differences in data generated by conventional testing and in-store research.

"The In-Store research showed XXXXX's (attribute) as having a larger effect on XXXXX's acceptance (25% rejection) than any previous studies. Specifically, EBES results have consistently showed about 2-3% claiming they would not purchase XXXXX for (attribute) reasons. In five months of U&A's, 5% of those interviewed commented that they did not like XXXXX's (attribute). We believe these results reflect test method differences i.e. point-of-purchase vs. home telephone interviews. We also believe point-of-purchase results reflect more accurately why consumers do and do not purchase XXXXX."