Search

Routines shape many aspects of day-to-day consumption. While prior research has established the importance of habits in consumer behavior, little work has been done to understand the implications of routines—which the authors define as repeated behaviors with recurring, temporal structures—for customer management. One reason for this dearth is the difficulty of measuring routines from transaction data, particularly when routines vary substantially across customers. The authors propose a new approach for doing so, which they apply in the context of ridesharing. They model customer-level routines with Bayesian nonparametric Gaussian processes, leveraging a novel kernel that allows for flexible yet precise estimation of routines. These Gaussian processes are nested in inhomogeneous Poisson processes of usage, allowing the authors to estimate customers' routines and decompose their usage into routine and nonroutine parts. They show the value of detecting routines for customer relationship management in the context of ridesharing, where they find that routines are associated with higher future usage and activity rates, and more resilience to service failures. Moreover, the authors show how these outcomes vary by the types of routines customers have, and by whether trips are part of the customer's routine, suggesting a role for routines in segmentation and targeting. [ABSTRACT FROM AUTHOR]

Marketing research relies on individual-level estimates to understand the rich heterogeneity of consumers, firms, and products. While much of the literature focuses on capturing static cross-sectional heterogeneity, little research has been done on modeling dynamic heterogeneity, or the heterogeneous evolution of individual-level model parameters. In this work, the authors propose a novel framework for capturing the dynamics of heterogeneity, using individual-level, latent, Bayesian nonparametric Gaussian processes. Similar to standard heterogeneity specifications, this Gaussian process dynamic heterogeneity (GPDH) specification models individual-level parameters as flexible variations around population-level trends, allowing for sharing of statistical information both across individuals and within individuals over time. This hierarchical structure provides precise individual-level insights regarding parameter dynamics. The authors show that GPDH nests existing heterogeneity specifications and that not flexibly capturing individual-level dynamics may result in biased parameter estimates. Substantively, they apply GPDH to understand preference dynamics and to model the evolution of online reviews. Across both applications, they find robust evidence of dynamic heterogeneity and illustrate GPDH's rich managerial insights, with implications for targeting, pricing, and market structure analysis. [ABSTRACT FROM AUTHOR]

Marketing managers are responsible for understanding and predicting customer purchasing activity. This task is complicated by a lack of knowledge of all of the calendar time events that influence purchase timing. Yet, isolating calendar time variability from the natural ebb and flow of purchasing is important for accurately assessing the influence of calendar time shocks to the spending process, and for uncovering the customer-level purchasing patterns that robustly predict future spending. A comprehensive understanding of purchasing dynamics therefore requires a model that flexibly integrates known and unknown calendar time determinants of purchasing with individual-level predictors such as interpurchase time, customer lifetime, and number of past purchases. In this paper, we develop a Bayesian nonparametric framework based on Gaussian process priors, which integrates these two sets of predictors by modeling both through latent functions that jointly determine purchase propensity. The estimates of these latent functions yield a visual representation of purchasing dynamics, which we call the model-based dashboard, that provides a nuanced decomposition of spending patterns. We show the utility of this framework through an application to purchasing in free-to-play mobile video games. Moreover, we show that in forecasting future spending, our model outperforms existing benchmarks. [ABSTRACT FROM AUTHOR]