David A. Schweidel

Online Product Opinions: Incidence, Evaluation, and Evolution

Whereas recent research has demonstrated the impact of online product ratings and reviews on product sales, we still have a limited understanding of the individuals decision to contribute these opinions. In this research, we empirically model the individuals decision to provide a product rating and investigate factors that influence this decision. Specifically, we consider how previously posted ratings may affect an individuals posting behavior in terms of whether to contribute (incidence) and what to contribute (evaluation), and we identify selection effects that influence the incidence decision and adjustment effects that influence the evaluation decision. n nAcross individuals, our results show that positive ratings environments increase posting incidence, whereas negative ratings environments discourage posting. Our results also indicate important differences across individuals in how they respond to previously posted ratings, with less frequent posters exhibiting bandwagon behavior and more active customers revealing differentiation behavior. These dynamics affect the evolution of online product opinions. Through simulations, we illustrate how the evolution of posted product opinions is shaped by the underlying customer base and show that customer bases with the same median opinion may evolve in substantially different ways because of the presence of a core group of “activists” posting increasingly negative opinions.

Listening In on Social Media: A Joint Model of Sentiment and Venue Format Choice

In this research, the authors jointly model the sentiment expressed in social media posts and the venue format to which it was posted as two interrelated processes in an effort to provide a measure of underlying brand sentiment. Using social media data from firms in two distinct industries, they allow the content of the post and the underlying sentiment toward the brand to affect both processes. The results show that the inferences marketing researchers obtain from monitoring social media are dependent on where they “listen” and that common approaches that either focus on a single social media venue or ignore differences across venues in aggregated data can lead to misleading brand sentiment metrics. The authors validate the approach by comparing their model-based measure of brand sentiment with performance measures obtained from external data sets (stock prices for both brands and an offline brand-tracking study for one brand). They find that their measure of sentiment serves as a leading indicator of the changes observed in these external data sources and outperforms other social media metrics currently used.

A Bivariate Timing Model of Customer Acquisition and Retention

Two widely recognized components, central to the calculation of customer value, are acquisition and retention propensities. However, while extant research has incorporated such components into different types of models, limited work has investigated the kinds of associations that may exist between them. In this research, we focus on the relationship between a prospective customers time until acquisition of a particular service and the subsequent duration for which he retains it, and examine the implications of this relationship on the value of prospects and customers. n nTo accomplish these tasks, we use a bivariate timing model to capture the relationship between acquisition and retention. Using a split-hazard model, we link the acquisition and retention processes in two distinct yet complementary ways. First, we use the Sarmonov family of bivariate distributions to allow for correlations in the observed acquisition and retention times within a customer; next, we allow for differences across customers using latent classes for the parameters that govern the two processes. We then demonstrate how the proposed methodology can be used to calculate the discounted expected value of a subscription based on the time of acquisition, and discuss possible applications of the modeling framework to problems such as customer targeting and resource allocation.

Portfolio Dynamics for Customers of a Multiservice Provider

Multiservice providers, such as telecommunication and financial service companies, can benefit from understanding how customers service portfolios evolve over the course of their relationships. This can provide guidance for managerial issues such as customer valuation and predicting customers future behavior, whether it is acquiring additional services, selectively dropping current services, or ending the relationship entirely. In this research, we develop a dynamic hidden Markov model to identify latent states that govern customers affinity for the available services through which customers evolve. In addition, we incorporate and demonstrate the importance of separating two other sources of dynamics: portfolio inertia and service stickiness. We then examine the relationship between state membership and managerially relevant metrics, including customers propensities for acquiring additional services or terminating the relationship, and customer lifetime value. Through a series of illustrative vignettes, we show that customers who have discarded a particular service may have an increased risk of canceling all services in the near future (as intuition would suggest) but also may be more prone to acquire more services, a provocative finding of interest to service providers. Our findings also emphasize the need to look beyond the previous period, as in much current research, and consider how customers have evolved over their entire relationship in order to predict their future actions. n nThis paper was accepted by Pradeep Chintagunta, marketing.

Incorporating Direct Marketing Activity into Latent Attrition Models

When defection is unobserved, latent attrition models provide useful insights about customer behavior and accurate forecasts of customer value. Yet extant models ignore direct marketing efforts. Response models incorporate the effects of direct marketing, but because they ignore latent attrition, they may lead firms to waste resources on inactive customers. n nWe propose a parsimonious model that allows direct marketing to impact three relevant behaviors in latent attrition models---the frequency with which customers conduct transactions, the size of the transactions, and the duration for which customers remain active. Our model also accounts for how the organization targets its direct marketing across individuals and over time. n nUsing donation data from a nonprofit organization, we find that direct marketing increases donation incidence for active donors. However, our analysis also shows that direct marketing has the potential to shorten the length of a donors relationship. We find that our proposed model offers superior predictive performance compared with models that ignore the impact of direct marketing activity or latent attrition. We demonstrate the managerial applicability of our modeling approach by estimating the impact of direct marketing on donation behavior and identifying those donors most likely to conduct transactions in the future.

Modeling Customer Lifetimes with Multiple Causes of Churn

Customer retention and customer churn are key metrics of interest to marketers, but little attention has been placed on linking the different reasons for which customers churn to their value to a contractual service provider. In this paper, we put forth a hierarchical competing-risk model to jointly model when customers choose to terminate their service and why. Some of these reasons for churn can be influenced by the firm e.g., service problems or price--value trade-offs, but others are uncontrollable e.g., customer relocation and death. Using this framework, we demonstrate that the impact of a firms efforts to reduce customer churn for controllable reasons is mitigated by the prevalence of uncontrollable ones, resulting in a “damper effect” on the return from a firms retention marketing efforts. We use data from a provider of land-based telecommunication services to demonstrate how the competing-risk model can be used to derive a measure of the incremental customer value that a firm can expect to accrue through its efforts to delay churn, taking this damper effect into account. In addition to varying across customers based on geodemographic information, the magnitude of the damper effect depends on a customers tenure to date. We discuss how our framework can be used to tailor the firms retention strategy to individual customers, both in terms of which customers to target and when retention efforts should be deployed.

Binge Watching and Advertising

How users consume media has shifted dramatically as viewers migrate from traditional broadcast channels toward online channels. Rather than following the schedule dictated by television networks and consuming one episode of a series each week, many viewers now engage in binge watching, which involves consuming several episodes of the same series in a condensed period of time. In this research, the authors decompose users’ viewing behavior into (1) whether the user continues the viewing session after each episode viewed, (2) whether the next episode viewed is from the same or a different series, and (3) the time elapsed between sessions. Applying this modeling framework to data provided by Hulu.com, a popular online provider of broadcast and cable television shows, the authors examine the drivers of binge watching behavior, distinguishing between user-level traits and states determined by previously viewed content. The authors simultaneously investigate users’ response to advertisements. Many online video providers support their services with advertising revenue; thus, understanding how users respond to advertisements and how advertising affects subsequent viewing is of paramount importance to both advertisers and online video providers. The results of the study reveal that advertising responsiveness differs between bingers and nonbingers and that it changes over the course of online viewing sessions. The authors discuss the implications of their results for advertisers and online video platforms.

A Multiactivity Latent Attrition Model for Customer Base Analysis

Customer base analysis is a key element in customer valuation and can provide guidance for decisions such as resource allocation. Yet extant models often focus on a single activity, such as purchases from a retailer or donations to a nonprofit organization. These models do not consider other ways that an individual may engage with an organization, such as purchasing in multiple brands or contributing user-generated content. In this research, we propose a framework to generalize extant models for customer base analysis to multiple activities. n nUsing the data from a website that allows users to purchase digital content and/or post digital content at no charge, we develop a flexible “buy til you die” model to empirically examine how the two activities are related. Compared with benchmarks, our model more accurately forecasts the future behavior for both types of activities. In addition to finding evidence of coincidence between the activities while customers are “alive,” we find that the latent attrition processes are related. This suggests that conducting one type of activity is informative of whether customers are still alive to conduct another type of activity and, consequently, affects inferences of customer value.

Social Media Intelligence: Measuring Brand Sentiment from Online Conversations

With the proliferation of social media, questions have begun to emerge about its role in providing marketing insights. In this research, we investigate the potential to “listen in” on social media conversations as a means of inferring brand sentiment. Our analysis employs data collected from multiple website domains, spanning a variety of online venue formats to which social media comments may be contributed. We demonstrate how factors relating to the focus of social media comments and the venue to which they have been contributed need to be explicitly modeled when deriving measures of online brand sentiment. Thus, we propose a model that separates the underlying brand sentiment from the effects of other predictable factors on social media comments. We apply our model to data pertaining to a leading enterprise software brand and show how our proposed approach provides an adjusted brand sentiment metric that is correlated with the results of an offline brand tracking survey. In contrast, a simple average of sentiment across all social media comments is uncorrelated with the same offline tracking survey. We also apply our modeling framework to social media comments related to three brands in different industries. From these analyses, we further demonstrate the potential pitfalls associated with simple average sentiment measures. We conclude by discussing the implications of our findings for practitioners who are considering social media as a potential research tool.

Dynamic Changepoints Revisited: An Evolving Process Model of New Product Sales

This paper posits a new framework to model the trial of and repeat purchasing of a new product. While much research has examined underlying shifts in consumer purchasing patterns, the typical assumption has been that the underlying purchasing process remains the same although the purchasing rates may change over time. Motivated by Fader, Hardie, and Huangs development of a dynamic changepoint model (2004), we consider an evolving process as consumers gain more experience with a new product.Our framework assumes that consumers progress through two purchasing states, becoming more regular in their inter-purchase times as they gain experience with the product through repeat purchases. More specifically, they move from an initial state of exponential purchasing to a steady state that is characterized by a more regular Erlang-2 timing distribution. The proposed model is very flexible and nests a number of existing models, enabling it to explain a wide range of observed behavioral patterns. We apply our evolving process model to the same datasets used by Fader, Hardie, and Huang (2004) and compare our results to a number of competing models. We find empirical evidence to support the use of a two-state model, since it yields relevant insights as well as improved empirical performance. We discuss the implications as they relate to forecasting new product sales.