An article I read some time ago by UC Berkeley professors David Aaker and Gary Shansby had some very interesting points that I would like to relate to a recent project we undertook. The customer was an auto dealership and wanted to know how his product was segmented and perceived from other products. So what this company did was decide to find out what these perceptions were by using touch screen displays installed in their various dealerships. The feedback was great. The displays asked questions something like the following…

“With respect to its class I would consider the [car make and model] to be:

• Sporty
• Roomy
• Economical
• Good Handling

“I would expect the typical [car make and model] owner to be:

• Older
• Wealthy
• Independent
• Intelligent

“The [car make and model] is most appropriate for:

• Short neighborhood trips
• Commuting
• Cross-country sight seeing

A host of other similar questions were asked of visitors to the dealership. The dealership had the questions online, but they found the response rate in the store–especially when it was attached to the prize giveaway of a new car–was very enlightening.

Some might be asking themselves “why on earth is he bringing this up?” It reminded me of an article I read sometime ago which spoke of multidimensional scaling. There were essentially two types: product-association-based and similarities-based.

Product-association-based scaling was obtained by a simple sampling of the target segment to scale the various objects on the product associations dimensions. The potential customer is simply asked how well they agree or disagree with statements regarding a particular product.

Similarly, the similarities-based multidimensional scaling “measures simply reflect the perceived similarity of two objects. For example, respondents may be asked to reate the degree of similarity of assorted object paris without a product association list which implicitly suggests criteria to be included or excluded. The result, when averaged over all respondents is a similarity rating for each object pair.”

The data gained from these type of analysis are then used to analyzes customers of various products and their perceptions. They answer questions like, “how is the market segmented? What really motivates the customer? What habits and behavior patters are relevant? What role does the product class play in purchase decisions? etc.

While in this particular setting the testing was used for potential purchasers of automobiles, this type of analysis could be used in a myriad of retail settings as a customer or clientele feedback device. This is also another way to measure feedback metrics of your display network, giving a pilot room for potential expansion. The questions I started asking myself after rereading through the study of the UC Berkeley professors was, “how do people perceive digital signage and signage software?”