There are many reasons why marketers want to go beyond an understanding of what consumers do and also understand why they do it. It is particularly useful to understand 1) what product and brand attributes drive customer purchase and loyalty (product performance, reliability, brand status, price, etc.) and 2) how do consumers rank the products of your company and your competitors on these characteristics? Clearly, an understanding of these factors would allow companies to:

  • Communicate with their customers in ways that emphasize the most valued attributes;
  • Target the weaknesses of competitive products;
  • Design new products that better meet the needs of customers;
  • Target different products towards the customers that most value their attributes.

It is sometimes possible to infer motivations from customer behavior. But in many cases the use of traditional market research (i.e., surveys) can provide greater insight into the motivation of customers. There are two specific types of market research that I will discuss: attribute importance and performance scores, and trade-off analysis.
Attribute importance and performance scores are obtained using straightforward consumer surveys. Respondents are asked about a list of product or brand attributes and provide feedback about the importance of each attribute in their buying decision and their perception of how each product or brand performs against these attributes. Typical results of this research are shown Figure 1.

Figure 1: Research from Consumer Surveys

This product is viewed as highly reliable, useful, fun and innovative. However, it is viewed as very high-priced, does not enhance status and has poor service support. Given the low importance given to service, it may not pay to invest in improving this score. However, given the strong scores in product functionality there seems to be a possibility of improving the product status through advertising. While the price is high, strong scores for other attributes may allow the company to maintain a high price.

While this analysis provides a good snapshot of attribute importance and performance, it is possible to use modeling to get much more specific about how variables are related. The conceptual layout of a possible structural equation model is shown in Figure 2.

Figure 2: Conceptual Layout of Structural Equation Model

First, some explanation of the model. It is assumed that the true underlying factors that drive purchase intent – and even purchase intent itself – are not observable. Instead, survey responses are used as indicators of the underlying variables. Observed purchase behavior is also used as an indicator of purchase intention. These underlying factors are influenced by other observable variables. (Of course, these relationships are not exact. But I have omitted the random elements of the model for simplicity.)

What information could this model provide beyond what was learned directly from the survey responses? First, it quantifies the size of the influence of each purchase driver on the actual probability of purchase. The survey captured responses on attribute importance, but could not link these directly to purchase intent. Second, by using actual purchase behavior as one indicator of purchase intent, it combines information from surveys and transaction data. This provides a more powerful overall model. Third, it quantifies the variables that influence the purchase process. For example, is price perception driven mostly by price advertising or by the actual price differential?

The most commonly used type of trade-off analysis is conjoint analysis. This is another way to combine consumer survey responses with a model in order to understand the underlying drivers of purchase. In this case, the survey consists of providing the respondents with a series of choices about product attribute “packages.” Figure 3 shows the choices might be offered.

Figure 3: Respondents have a Series of Choices

For each set of choices, the respondent would say which product is preferred. The choices are designed in such a way that the underlying value the respondent places on each attribute can be determined. By summing the results across respondents, it is possible to determine the overall demand impact of changing any attribute.

Tremendous marketing advantage can be gained from data mining in huge transactional data sets. But it is worth remembering that traditional research and modeling can also add value.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access