Summary
In the context of key driver analysis in applied customer satisfaction research, the assumption of sample homogeneity (that single models perform adequately over the entirety of a survey sample) can be shown to restrict the value of the insights derived. While latent class regression has been used as a method of circumventing some of these issues, it is proposed that there are major barriers to both uptake and successful practical usage of the technique. Several of these issues are common to any multivariate technique, while others are specific to latent class regression. Following an examination of these issues, we introduce an alternative technique for deriving discrete latent classes, using a combination of genetic algorithms and (bivariate) correlations. This paper concludes that the proposed approach outperforms latent class regression in its ability to deliver action-orientated insights, and is better placed to assist marketers facing real-world research questions and datasets.

Summary
This paper reviews the current state of use (and value) of quantitative research in advertising. It concludes that the value of quantitative research is well established but use is less than might be expected based on value. The authors discuss the reasons for this difference in value.

Summary
The purpose of this paper is to offer a better understanding of the effects of product-harm crisis on a brand’s performance and market structure. This research is based on panel data on milk powder sales during the Nestlé product-harm crisis in China. The NBD-Dirichlet model is used to evaluate the performance of Nestlé and other leading milk powder brands before, during and after the crisis. Our data show that product-harm crises disturb the market structure and change customer behaviour. While a product-harm crisis had a negative effect on Nestlé’s brand performance, it created opportunities for other brands. Overall, our analysis shows that the NBD-Dirichlet model is a valid tool for monitoring the performance changes of both crisis brand and other non-crisis brands during a product-harm crisis. The managerial implications are also discussed.

Summary
The effectiveness of three commonly used methodologies for market size estimation, as well as their accuracy in explaining the future behavior of the automobile market in Ecuador, is addressed in this presentation. These methodologies are Delphi forecast based on local expertise, Cross sectional linear regression, and Time series modeling of sales. Data of car sales ranging between 1990 and 2009 was used to compare the performance of these three methodologies. We found that Experto Forecast performs very well in the short term, while Time Series Modelling fits better the behavior of the market for medium term predictions.

Summary
Discrete choice experiments are analysed using multinomial logit models. One key trait of these models is that independent variables are usually based on alternative related characteristics, such as the price of different options or the commute time for different travel alternatives. Respondent level characteristics, or demographics, are not typically included as independent variables or moderating constructs since these parameters do not vary across options in a choice set and, as such, do not impact the corresponding choice probabilities. To address this weakness, the objective of the current paper is to share a practical and usable approach to incorporate demographic variables as moderating constructs in discrete choice experiments and multinomial logit models. This approach requires the computation of a new variable representing the interaction between the focal demographic variable and an alternative related characteristic. For illustrative purposes, this procedure is applied to hypothetical transportation data.

Summary
Brand equity is loosely defined as: 'the added value that a brand brings to a product or service beyond the functional benefits provided'. Structural equations modelling (SEM) is a sophisticated way of measuring brand equity that gives lucid results that are easy for marketers to interpret. The chief strength of SEM is that it can be expressed in path diagrams, allowing clients and marketing managers to understand the output of a structural equations model with a minimum of explanation. Its other main advantage over other multivariate methods is that it includes the ability to handle latent variables in the analysis.

Summary
The most successful brand extensions are considered to be those having high perceived similarity between the parent brand and the extensions, and being well known in the marketplace. However, previous research has mainly examined the effects of overall measures of perceived similarity between a parent brand and an extension. Correspondingly, little is known about the effects of different areas of consumer knowledge. This study investigates the effects of three types of perceived similarity (usage, associations, competence) and three areas of consumer knowledge (original brand, original category, extension category) on evaluations of brand extensions. The results indicate that some types of perceived similarity and knowledge are more important than others. These findings imply that brand managers need to identify and measure the relevant types of perceived similarity and knowledge that will affect evaluations of brand extensions in order to design effective communication strategies for extensions.

Summary
The practice of brand tracking research has changed little in the past 30 years. Methodological advances in choice modelling, not typically applied to brand tracking, represent a potentially valuable enhancement of brand tracking practice. A case study of frozen pizza brand choice illustrates the possibilities, with particular attention paid to the MNL and lexicographic models.

Summary
How can the importance of consumer wants and needs be quantified? Using data sets from multiple consumer brand and advertising tracking studies, several standard traditional market research techniques are compared to vector autoregression (VAR) modeling. It is demonstrated that by utilizing VAR models and resolving causal ambiguity, key performance indicators can be identified that not only correlate with traditional market research summary metrics, such as overall ratings and purchase interest, but that also drive brand sales/share and thereby qualify as metrics that matter. The analytic philosophy underlying the VAR analytic approach also is shown to be consistent with (and complementary to) market mix modeling analysis. Presented is a procedure for the simultaneous assessment of the relative and absolute impact of multiple marketing initiatives on baseline and incremental sales—including advertising and promotion effects and traditional consumer awareness and attitudinal metrics—facilitating resource-allocation decisions and providing marketers within a single framework for return on marketing investment optimization.

Summary
Marketing managers typically have to use and integrate many pieces of data and marketing intelligence when taking decisions such as whether to launch a product and, if so, at what price. Conjoint experiments and analysis remain popular marketing research tools with business practitioners to test and measure how the market will react to different actions. There is a growing body of work that focuses on, first, how to construct the experiments so that they better represent real market conditions and, second, the use of sophisticated model specifications that provide information on consumers’ responses. The market researcher typically uses internal validation for model validity – a comparison of model prediction and within-sample holdout data. We contend in this paper that customers and users of market research information need to adopt a different and wider meaning of validity, referred to as external validity, to facilitate improved decision making. In this research, a case study is used as an example to demonstrate how marketing managers can use the information from a choice-based conjoint derived choice model differently depending on the manner in which the model validation is carried out.