Invited Paper—Learning Models: An Assessment of Progress, Challenges, and New Developments

Abstract

Learning models extend the traditional discrete choice framework by postulating that consumers have incomplete information about product attributes, and that they learn about these attributes over time. In this survey we describe the literature on learning models that has developed over the past 20 years, using the model of Erdem and Keane (1996) as a unifying framework. We described how subsequent work has extended their modeling framework, and applied learning models to a wide range of different products and markets. We argue that learning models have contributed greatly to our understanding of consumer behavior, in particular in enhancing our understanding of brand loyalty and long run advertising effects. We also discuss the limitations of existing learning models and discuss potential extensions. One key challenge is to disentangle learning as a source of dynamics from other key mechanisms that may generate choice dynamics (inventories, habit persistence, etc.). Another is to enhance identification of learning models by collecting and utilizing direct measures of signals, perceptions and expectations. 1 Until fairly recently, the logit was much more popular than probit, largely due to computational advantages. But advances in simulation methods, such as the GHK algorithm and Gibbs sampling (see Geweke and Keane (2001), McCulloch, Polson and Rossi (2000)) have greatly increased the popularity of probit, particularly among Bayesians. 2 Thus, learning models play havoc with traditional welfare analysis, as consumer surplus is no longer the area under the demand curve. Parameters of the demand curve are no longer structural parameters of preferences, but depend on the information set (e.g., they can be shifted by advertising). See Erdem, Keane and Sun (2008) for a discussion. 2 exogenously, or "active search," exerting effort to gather information. Or they may do both.A fourth distinction is how consumers learn. They may be Bayesians, or they may update perceptions in some other way. For instance, consumers may over/under weight new information relative to an optimal Bayesian rule, or forget information that was received too far in the past.Learning models were first applied to marketing problems in pioneering work by Roberts and Urban (1988) and Eckstein, Horsky and Raban (1988). 3 But, due to technical limitations of the time (both in computer speed and estimation algorithms), their models had to be quite simple. Roberts and Urban (1988) study how Bayesian consumers learn about a new product from wordof-mouth signals. Consumers in their model are risk averse, but myopic, so there is no active search. In contrast, in Eckstein et al (1988) consumers are forward-looking, and trial purchase is the (only) source of information. But utility is linear, so their model exhibits the "value of information" phenomenon, but not the "brand equity" phenomenon created by risk aversion. For Roberts and Urban (1988) the converse is true. The strong simplifying assumptions of these early models, plus the difficulty of estimating...

Bayesian Revision vs. Information Distortion

“…More specifically, their use by market researchers relies on both past consumption and internet product search to predict future consumption and, more recently, further search (e.g., Ching et al, 2013; Fong, 2017). However, even as these “consumer learning” models become more sophisticated, they tend to be revised between purchases only by the knowledge of what consumer has searched.…”

An Empirical Study of Customer Spillover Learning About Service Quality

“…Our paper uses a Bayesian learning modeling framework. Ching et al (2013) provide an extensive literature review and identifies four categories of recent developments. Our work falls into the category: models of correlated learning.…”

“…Our work falls into the category: models of correlated learning. Ching et al (2013) summarize the few existing studies in this stream (e.g., Erdem 1998, Coscelli and Shum 2004, Sridhar et al 2012, Chan et al 2013) and indicates that correlated learning means "· · · learning about a brand in one category by using the same brand in another category and/or learning about one attribute (e.g., drug potency) from another (e.g., side effects). This occurs if priors and/or signals are correlated across products or attributes · · · ".…”

“…This uncertainty makes the standard random utility framework unsuitable because it typically assumes that customers know the attributes of all alternatives perfectly before making a choice, such as in Aksin et al (2013). In contrast, a learning model is more suitable to our context (e.g., Ching et al 2013), because they allow customers to have incomplete information about product attributes. Specifically, customers form expectations about these attributes from one or several information sources such as past experiences with the product, word of mouth and advertisements, and then use their beliefs about service performance in their purchase decisions.…”

Learning to Trust Flu Shots: Quasi‐Experimental Evidence from the 2009 Swine Flu Pandemic

“…Consumers are often uncertain about the benefits and risks of medical products and treatments, including vaccines, which are generally regarded as credence or experience goods (see, e.g., Pita Barros and Martinez Giralt, ). Consumer learning describes a process in which personal consumption experiences result in new insights, attitudes and beliefs regarding uncertain product characteristics, which in turn provide the basis for future product demand (Ching et al, ). Consumer learning may thus be an important process determining demand for health products with uncertain benefits and risks, such as influenza vaccine.…”