Getting the right information is essential to making the right decisions. Good research can help you make the right decisions by providing you with the right information.....

Key Driver Analysis:

With Key Driver Analyses, the focus is placed on identifying the “key drivers” of satisfaction, loyalty, or retention. The goal of this type of analysis is to identify the smallest set of attributes (performance/features/attributes) that has the greatest capacity to alter/improve overall satisfaction, loyalty, or retention – those elements that are most important in driving satisfaction/loyalty/retention. In general, we use two broad categories of methods are used:

• Indirect Methods – “back room” statistical and/or mathematical methods used to identify key drivers. Three of the more common methods used to identify underlying relationships and relative importance are correlations, factor analysis, and multiple regression. These types of methods can often be easily used with existing CSM databases and are minimally intrusive. However, the results are limited to the quality/depth of the information available as well as the multicollinearity of the data.
• Direct Methods – methods used to identify key drivers directly from customers in a separate research initiative. Some of the more common methods used to identify relative importance are importance ratings, trade-off methods (rank ordering, constant sum, and pair preference), and card sorts. These types of methods require separate/additional research, but greatly maximize the likelihood of discovering new issues because they are not limited by the information currently in-house.

Satisfier/Dissatisfier Analysis™:

The Satisfier/Dissatisfier Profile™ is a method of analyzing customer satisfaction, loyalty, or retention information developed by Bosma Research International. The Profile reveals the extent to which meeting or not meeting customer expectations across specific performance attributes affects overall satisfaction, loyalty, or retention. In particular, the Profile identifies satisfiers (things that, if done well, create satisfaction, loyalty, or retention) and dissatisfiers (things that, if done poorly, create dissatisfaction, loyalty, or retention).

The Profile can best be described as a composite representation of the specific aspects that promote satisfaction/loyalty/retention as well as the aspects that tend to arouse dissatisfaction/disloyalty/defection. The power of the Profile can be viewed in the following two examples.

The Results of a Traditional Key Driver Analysis:

The traditional key driver analysis can identify the relative importance of Scale of Operations, Aircraft, In-flight Service, and On-Time Performance as key drivers of overall customer satisfaction for this market segment.

The Results of the Same Scenario Using a Satisfier/Dissatisfier Analysis™:

An analysis of the same scenario using a Satisfier/Dissatisfier Analysis™ reveals that with this particular market segment, the negative impact and/or consequences of not meeting customers’ expectations across these four service areas is substantially greater than the potential increase to be gained. It becomes very clear that there is little room for error across any of these four service areas with this customer group. As a result, a service priority for the airline would likely need to be doing things right the first time with this customer group.

The Results of the Same Satisfier/Dissatisfier Analysis™ for a Different Segment:

This second example presents both the power and capacity of the Profile to identify clear differences in service/product expectations across different market segments.

Structural Equation Modeling (SEM):

SEM is an analytic procedure used to test the directional relationship among variables in a customer satisfaction, loyalty, or retention model. The first step in the process is to develop a structural (conceptual) model that best represents your understanding of the underlying relationships and relative importance of the different variables. Factor analysis and multiple regression can be used to help you develop your model. The second step is then to test/examine the statistical relationship among the variables using statistical procedures such as LISREL or AMOS.

Developing A Structural Model – The structural model represents your hypothesis about the relationships among the individual attributes/variables (performance/features/attributes) as well as the sets of scales/constructs that have evolved from combinations of the attributes. A structural model might be constructed as follows:

Building The Measurement Model – The second step is then to test/examine the statistical relationship among the variables using statistical procedures such as LISREL or AMOS. It is during this phase of SEM that the statistical/mathematic relationships among individual attributes/variables (performance/features/attributes) and the sets of scales/constructs developed are generated. SEM can provide insights into the relationships that exist within a model:

• variables that are related with neither one having an impact on the other one or another
• variables that are related in which one of the variables is an influencer of the other

Market Segmentation:

Market segmentation is about defining markets and competitive environments in ways that are relevant for understanding and influencing the competitive behavior of customers. It is the process of categorizing customers (or potential customers) into segments that are maximally homogeneous within each group while being maximally heterogeneous between groups and then deciding which group(s) offer the greatest opportunity for a given product/service.

Many businesses today spend a great deal of time and effort conducting segmentation and customer satisfaction studies as a means of understanding current/potential customer behavior, targeting service/product delivery “hot spots”, and building customer loyalty. All too often, segmentation and customer satisfaction measurement activities are conducted independently of one another, and they often fail to create a clear linkage between customer loyalty strategies and organizational effectiveness, bottom-line, and process effectiveness measures. A comprehensive customer measurement system needs to have all of these components integrated.

By definition market segmentation is the process of categorizing consumers into groups or “segments” that are maximally homogeneous within each group (based on some unique set of distinguishing characteristics) while, at the same time, being maximally heterogeneous between groups.

In general, there are two approaches/types to segmentation:

• Descriptive Segmentation – segmentation conducted to identify and/or describe the differences between different groups of consumers – provide little or no basis for understanding customer behavior or changing choice behavior.
• Predictive Segmentation – segmentation conducted to provide understanding of the differences between different groups of consumers and through that understanding provide a means of predicting and changing choice behavior.

Trade-Off Analysis:

Trade-off analysis allows you to study the impact of product/service features, pricing, packaging, and competitive changes all at the same time. Begun as a technique named Conjoint Analysis back in the 70's, trade-off analysis is a family of marketing research techniques that can tell you the collective set of product/service features that can be built into a new product to generate the most sales, or which price will maximize your gross profit and market share, or who will be your core customer for this new product, or how many you can expect to sell in the first year.

In essence, trade-off analysis allows the researcher to throw all of his/her options into a carefully constructed questionnaire. Respondents are then asked a series of product purchase intent questions. The data are then subjected to a variety of advanced statistical procedures which are then used to create mathematical models – these models allow decision makers to simulate the marketplace in great detail and with surprising accuracy.

Trade-off models can be very useful and allow us to answer a variety of questions, some of which include:

• Which product/service features and/or combinations of product/service features should we include in the final product?
• What price should we charge? Who should we be selling the product/service to?
• To what extent do customers make trade-offs between product, price, service, and brand attributes? What is the geographic distribution of likely customers?
• How can we increase/establish market share? How can we ensure that our current customers stay our customers?

In the context of a new product development study, trade-off analysis can also be used to help develop an overall strategy for introducing a new produce/service to the marketplace. Trade-off models can be used to obtain answers to following types of questions:

• Which of the potential product/service configurations should we pursue?
• What value propositions should be offered for a given product/service configuration?
• What positioning should be used for each product/service configuration for a new entrant relative to its competitors?
• Where should a new company deploy first?
• What marketing tactics should a new entrant to the marketplace pursue to maximize market share and presence?
• What target markets should the new entrant pursue?