Data pruning in consumer choice models

Common, if not ubiquitous, Marketing practice when estimating models for scanner panel data is to: (a) observe the data, (b) prune the data to a “manageable” number of brands or SKUs, and (c) fit models to the remaining data. We demonstrate that such pruning practice can lead to significantly different (and potentially biased) elasticities, and hence different managerial/practical outcomes, especially in the context of model misspecification. We first justify our claims theoretically by writing the general problem in a classic missing-data framework and demonstrate that commonly used pruning mechanisms (gleaned from current academic Marketing literature) can lead to a nonignorable missing data mechanism. Secondly, we summarize an extensive set of simulations that were run to understand the driving factors of that bias. The results indicate much greater pruning bias in those cases where model fit is poor (small ), random utility errors are correlated with the covariates, or the model is misspecified (e.g., a homogeneous logit is specified when a mixed-logit is true). Empirically, we also demonstrate our findings on the well-cited and highly utilized fabric softener data of Fader and Hardie (1996). Our empirical findings suggest a number of estimates that vary according to the way in which the data is pruned including the magnitude of market mix and attribute elasticities, and purchase probabilities, but that the pruning effect is smaller for better fitting models.

Multi-tiered private labels portfolio strategies: Effects on consumer behavior

Abstract

Retailers wish to expand their standard private label (PL) adopting a multi-tiered portfolio. We study the effects produced by the introduction of two new PLs quality-tiers (economy and premium) on the market share of various national brands (NB; premium-quality and second-tier) and the standard PL. This study proposes a model that accommodates three effects (similarity, attraction, and compromise) for understanding how the introduction of economy and premium PL may affect market incumbents. This study also analyzes the effects of introducing new PL quality-tiers for customer segments. Our results indicate that when economy and premium PL are introduced in the market, the choice probability of standard PL decreases, especially for high purchase quantity customers and PL loyal customers. In addition, introduction of a premium PL decreases the choice probability of second-tier NB and premium-quality NB, especially for PL loyal customers and high purchase quantity customers.     

Internet and distance channel use and European consumer complaint behavior

This exploratory study analyzes the effect of distance-shopping channel choice (i.e. the format first used by consumers in a retail transaction) on consumer complaint behavior, with a special focus on the Internet channel. It analyzes 2011 Eurobarometer data using a logistic regression model controlling for sociodemographic and country-specific variables to measure the impact of using the Internet, post/catalog, and telephone channels on complaint behavior. The results indicate that consumers using the Internet channel exhibit the greatest likelihood of complaint behavior when compared to consumers using traditional distance-shopping channels. Implications and suggestions for retailers are discussed, as are the contributions to the consumer complaint behavior literature.     

Theory of planned behavior in consumer behavior research: A systematic literature review

This study aimed to systematically review and categorize studies on consumer behavior based on theory of planned behavior (TPB), its core constructs, or extensions, and to provide directions for future research agenda. Scopus and the Web of Science were consulted for studies based on TPB, its parts, or extensions. The inclusion criteria were studies published in peer-reviewed journals, in English, and within the past decade (i.e., between 2012 and 2021). Graphical methods were used to visually display research findings. For the purpose of literature clustering, MAXQDA 2020 software was employed. In total, 118 scientific, peer-reviewed sources were included in the review. Two categories, five clusters, and seven subclusters emerged from the literature set. The results revealed a significant research tendency toward exploring consumer green behavior and consumer purchase intention of food products. The least-explored research themes were focused on consumer intention toward and purchase behavior of clothing, green vehicles, and green personal care products. The review confirmed the growing prevalence of TPB in consumer behavior research aimed at exploring factors preceding behavior.     

The effect of activity identity fusion on negative consumer behavior

Negative consumer behavior is an important research topic as it explores consumer behaviors that threaten a brand's image and financial stability. However, prior research offers conflicting findings on whether a strong consumer‐brand relationship hurts or protects a brand after a market disruption or a brand transgression. To provide clarity on this issue, this work argues that disrupting the consumer‐activity relationship motivates consumers to reaffirm and protect their identity, thereby leading to negative consumer behavior. The data reveal that, after a brand‐initiated market disruption, consumers with high activity identity fusion are more likely to spread negative word‐of‐mouth, boycott the brand, and avoid repurchasing the brand in the future. Moreover, the data suggest that high brand identity fusion protects the brand during market disruptions; therefore, prior conflicting results may be due to the fact that the consumer‐activity relationship was not accounted for. Theoretically, this work establishes that consumers' relationship with their consumption activities has significant impacts on consumer behavior. Brand managers and marketers are urged to develop strategies that focus on strengthening the consumer‐brand relationship and not the consumer‐activity relationship.     

Incorporating choice dynamics in models of consumer behavior

This paper presents a framework for organizing and discussing factors influencing consumer choice dynamics, how these factors may be incorporated into models of buyer behavior and problems that may arise in estimating such models. The paper identifies research issues and delineates possible approaches. Proceedings of Session on Choice Dynamics at the Banff Symposium on Consumer Decision-Making and Choice Behavior. All authors share equally in content and remaining errors.     

The Impact of Price Disclosure on Dynamic Shopping Decisions

A potentially powerful way to assist consumers in making dynamic shopping decisions is to disclose price information to them before they shop, for example by posting prices on the Internet. This paper addresses the differential impact of disclosing either only current, or both current and future prices, on consumer shopping decisions in multi-period tasks involving multiple product purchases. In the context of an Internet-based experiment, we find that consumer expenditure deviates more strongly from that of a normative model when both current and future prices are disclosed than if only current prices are disclosed. We investigate the behavioral effects underlying this finding by estimating a model that allows for variations in consumer discounting, strength of store price format preferences, as well as choice consistency between different price disclosure conditions.     

Convergence and divergence in consumer behavior: implications for international retailing

Converging technology and disappearing income differences across countries will not lead to homogenization of consumer behavior. Rather, consumer behavior will become more heterogeneous because of cultural differences. As consumer incomes converge across countries, the manifestation of value differences will become stronger. This phenomenon makes it increasingly important to understand values of national cultures and their impact on consumer behavior. Retailing strategies for one country cannot be extended to other countries without adaptation. Hofstede’s model of national culture is applied to understand differences in consumer behavior across countries. Examples are provided of consumption differences, their relationships with culture discussed, and selected implications for international retailing management detailed.     

Retailer corporate social responsibility and consumer citizenship behavior: The mediating roles of perceived consumer effectiveness and consumer trust

Retailers often face pressure from stakeholders to address social and environmental issues. The purpose of this study is to investigate the influence of retailer corporate social responsibility (CSR) on consumers' perceptions and behavior. Drawing on the theoretical foundations of social identity theory and signaling theory, this study presents and validates a unique model which examines the mediating effects of perceived consumer effectiveness and consumer trust on the relationship between retailer CSR and consumer citizenship behavior. Using a sample of 407 consumers in China, results of structural equation modeling show that retailer CSR is positively associated with consumer citizenship behavior. Additionally, perceived consumer effectiveness and consumer trust positively mediate this association. Taken together, these findings provide support for retailers to develop CSR programs that remind and reinforce consumers’ perceptions about socially responsible behaviors.     

Consumer transactions with SMEs: implications for consumer scholars

Historically, the consumer movement has defined business as big business. This paper proposes that we shift the focus to consumer transactions with small‐ and medium‐sized enterprises (SMEs). After SMEs are defined, facts about the realities of SMEs are used as stepping stones to develop the idea that consumer interactions with SMEs are different than those with large firms and merit the attention of consumer scholars. Thirteen recommendations are tendered for future research including rethinking our traditional approach to consumer behaviour theory, consumer policy, and consumer movement theory.     

Estimating choice models in data-sparse environments: Taking advantage of perceived similarity

Research suggests that choice models conditioned on correctly identified consideration sets outperform choice models conditioned on the awareness set (Hauser, 1978; Roberts and Lattin, 1991). However, in data-sparse environments, where purchase history information is not available or not relevant, choice models conditioned on the consideration set often yield nonunique or nonsignificant solutions. In these environments, we propose the use of similarity information to improve the performance of choice models. Support for this position is found in an empirical application involving automobiles.     

Effects of voice assistant recommendations on consumer behavior

The present study compares the influence of text-based recommendations; traditionally known as online consumer reviews, and the influence of voice-based recommendations provided by voice-driven virtual assistants on consumer behaviors. Based on media richness theory, the research model investigates how voice versus text modality influences consumers' perceptions of credibility and usefulness, as well as their behavioral intentions and actual behaviors. In addition, the study analyses if these relationships vary based on the type of product and compares the influence of masculine and feminine voices. Two studies were conducted using between-subjects experimental designs, partial least squares-structural equation modeling, and logistic regression. The core finding is that voice-based recommendations are more effective than online consumer reviews in altering consumer behaviors. In addition, the first study showed that the influence of recommendations on behavioral intentions is mediated by consumers' perceptions of their credibility and usefulness. The second study confirmed, in a realistic setting, that voice-based recommendations affect consumer choices to a greater extent. Recommendations for search products and provided by males are also found to be more effective. These results contribute to the voice assistant and e-WOM literature by highlighting the effectiveness of voice-based recommendations in predicting consumer behaviors, confirming that credibility and usefulness are key factors that determine the influence of recommendations, and showing that recommendations are more effective when they focus on search products.     

Ambient scent and consumer behavior: a field study in a florist's retail shop

Numerous studies have shown that ambient scents influence human behavior including customers' behavior in real retail environments. However, the effects of scents in a closed-in plant and flower shop have not yet been examined. This study proceeded to measure customer behavior in two different situations taking place in a French flower shop: when lavender aroma was diffused and when it was not. The results showed that lavender aroma increased both the number of customers who bought plants and/or flowers and the amount of purchasing.     

Construction of consumer choice in the market: challenges for environmental policy

Drawing from the literature on the analytics of government, the paper discusses marketing as a form of government, elaborating and illustrating the many ways in which consumer choice is shaped, modified and directed in the market through practices and techniques of consumer marketing. The aim is to critically reflect upon and render problematic the individualistic ideas of the green consumer as a powerful market force and to provoke discussion on the conceptualization – and construction – of consumer subjectivity and social problems in marketing. Taking examples particularly from the fashion and clothing industry, the paper discusses the ways in which marketing activities come to shape consumer conduct by operating through the choice of individuals who freely pursue their needs and desires, and by working on the environment within which this freedom of choice is exercised. The paper contributes to the literature on green consumerism by systematically interrogating and elaborating on the modes and practices of marketing thought and expertise through which consumers and consumption are rendered intelligible and actionable in the market.     

The relative impact of corporate reputation on consumer choice: beyond a halo effect

ABSTRACT

Previous work suggests that corporate reputation generates a ‘halo effect’ where products from companies with better reputations are more likely to be chosen. We argue that corporate reputation plays a more expansive role, proposing that consumers will be less price-sensitive to offerings endorsed by companies with good reputations and that it moderates the marginal utility of product features with high clarity. We also propose that an individual’s knowledge of a company increases the likelihood its products will be purchased. Using a choice model incorporating an individual SEM-based reputation measure, we find support for these hypothesised effects in the context of television choices. The results suggest that corporate reputation warrants more attention by marketing managers to increase preferences for their products through these mechanisms.     

Hits and misses: Herd behavior and online product popularity

This article presents the results of a market experiment in which the perceived popularity of programs on a large commercial online system were manipulated. The download counts of software, which indicate how many previous users had obtained a software program, were artificially increased by repeated downloading. This manipulation was done to one member of a matched program pair, with the manipulation levels set randomly to achieve a balanced experimental design over a set of matched pairs. Subsequent downloads of the programs made by the service's subscribers were then recorded. Substantial additional downloads were induced by the manipulation. Some consumers apparently are drawn to market leaders, further increasing their lead. This raises an important issue regarding the need for verifiability of online popularity measures.     

Issues in the estimation and application of latent structure models of choice

Our paper provides a brief review and summary of issues and advances in the use of latent structure and other finite mixture models in the analysis of choice data. Focus is directed to three primary areas: (1) estimation and computational issues, (2) specification and interpretation issues, and (3) future research issues. We comment on what latent structure models have promised, what has been, to date, delivered, and what we should look forward to in the future.     

Experimental analysis of choice

Experimental choice analysis continues to attract academic and applied attention. We review what is known about the design, conduct, analysis, and use of data from choice experiments, and indicate gaps in current knowledge that should be addressed in future research. Design strategies consistent with probabilistic models of choice process and the parallels between choice experiments and real markets are considered. Additionally, we address the issues of reliability and validity. Progress has been made in accounting for differences in reliability, but more research is needed to determine which experiments and response procedures will consistently produce more reliable data for various problems.Sawtooth SoftwareDecision Research     

Neutralization techniques as a moderating mechanism: Ethically questionable behavior in the Romanian consumer context

On the basis of an empirical investigation in the context of Romania, this paper identifies a moderating role of neutralization techniques within ethically questionable consumer behavior. The quantitative study is based on a synthesized model of theory of planned behavior incorporating the factor of perceived unfairness and neutralization techniques. Significantly, neutralization techniques are shown to have a negative, but definite impact on the action to behave unethically. This leads to their consideration as a process of thinking, rather than as static judgment. As such, neutralization techniques are conceptually distinctive to the other factors. The paper analyzes the results specific to the Romanian context, but noting implications for an understanding of the morality of markets with similar historical, political, and economic conditions. Overall, the findings offer a more nuanced reading of consumer behavior. The paper places moral flexibility in terms of a specific cultural context, but also reveals how neutralization techniques can moderate ethically questionable behaviors beyond matters of self‐interest, which, in turn, has implications for how companies can consider their responsibilities in relation to their customers.     

Evaluating the potential of agent-based modelling to capture consumer grocery retail store choice behaviours

Evolving consumer behaviours with regards to store and channel choice, shopping frequency, shopping mission and spending heighten the need for robust spatial modelling tools for use within retail analytics. In this paper, we report on collaboration with a major UK grocery retailer to assess the feasibility of modelling consumer store choice behaviours at the level of the individual consumer. We benefit from very rare access to our collaborating retailers’ customer data which we use to develop a proof-of-concept agent-based model (ABM). Utilising our collaborating retailers’ loyalty card database, we extract key consumer behaviours in relation to shopping frequency, mission, store choice and spending. We build these observed behaviours into our ABM, based on a simplified urban environment, calibrated and validated against observed consumer data. Our ABM is able to capture key spatiotemporal drivers of consumer store choice behaviour at the individual level. Our findings could afford new opportunities for spatial modelling within the retail sector, enabling the complexity of consumer behaviours to be captured and simulated within a novel modelling framework. We reflect on further model development required for use in a commercial context for location-based decision-making.