Séminaires de recherche

Marketing Research Camp 2015


Intervenant : Thomas Otter, Goethe University, Frankfurt; Ashley Humphreys, Northwestern University; Janet Schwartz, Tulane University; and Shantanu Dutta, University of Southern California.

10 avril 2015 - Salle S228 - De 09h00 à 18h00

“Measuring Substitution and Complementarity among Offers in Menu Based Choice Experiments”
Tetyana Kosyakova, Goethe University, Frankfurt Thomas Otter, Goethe University, Frankfurt Christian Neuerburg, GfK, Nürnberg

Choice experiments designed to extend beyond the classic application of choice among perfect substitutes have become popular in marketing research. In these experiments, often referred to as menu based choice, respondents face choice sets that may comprise substitutes, complements, and offers that provide utility independently, or any mixture of these three types. The inferential challenge posed by data from such experiments is in the calibration of utility functions that accommodate a mix of substitutes, complements, and ''independent" offers. Moreover, while a prior understanding of the product categories under study may, for example, suggest that two offers in a set are essentially perfect substitutes, this may not be true for all respondents. To address these challenges, we combine the autologistic model with a flexible hierarchical prior structure. We explain from first principles how the autologistic model improves on the multivariate probit model and on models that assume independence but include cross-price effects in the utility function. We develop Bayesian inference for the autologistic model, including its intractable normalizing constant. Finally, we find empirical support for our model in a menu based conjoint experiment investigating demand for game consoles and accessories and we illustrate implications for optimal pricing.

"Professional Contests and the Emergence of Social Media as an Institutional Field"
Ashlee Humphreys, Northwestern University

This project investigates the legitimation of social media as an emerging professional field. Using qualitative and quantitative content analysis, we examine social media discourse from mainstream media in the United States, New York Times, Wall Street Journal, and USA Today (n=6,877) and trade journals such as AdAge, Ad Week, American Journalism Review, and PR News (n= 3,524). Drawing from Andrew Abbott’s work on the professionalization of fields, we find that social media professionals engaged in professional contests amongst marketing, journalism, and computer or information science. We show that social media emerged from a well-established field of ‘communication specialists,’ professionals working in marketing and advertising, but then evolved to incorporate more varied and deeper firm activities such as product development, long-term customer management, and broad public relations efforts. Techniques from computer science, such as machine learning and database management emerged, challenging some of the field’s boundaries, and eventually shaping social media activities toward the transactional, market-based activities. As such, the tasks of social media were initially conceptualized as communicative ones - reaching particular audiences, engaging them in product-centric communications for the purposes of brand building, but have transformed toward a greater emphasis on data collection and prediction.

The results of this study demonstrate the profound effects of professional contests on the activities of a broader consumer and media public. Following prior research on institutional logics and the professions, we show how social and cultural structures shape the paths of technological innovations as they enter fields of professional discourse and evaluate what this means for consumer life.

"Gain without Pain: The Extended Effects of a Behavioral Health Intervention"
By Janet Schwartz, Tulane University.

We examine the extended effects of a incentive-based behavioral health intervention that financially penalized shoppers who failed to improve the nutritional quality of their grocery basket. While the intervention successfully motivated the target behavior (Schwartz et al., 2014), less is known about any persistence and spillover effects that could lead to worse behavior in other domains or after the intervention ends. Because many theories predict that such balancing behaviors can occur, but rarely examine them in field experiments, this novel examination of the data presents an opportunity to advance our knowledge. Our results show that during the intervention the treatment group showed positive persistence effects after the penalty was removed, no negative spillover effects into related domains such as exercise during the intervention, and no negative impact on customer loyalty as a result of a financial penalty. These results highlight the importance of examining the extended effects of behavioral field interventions for both theory and practice.

"An Empirical Analysis of Market and Operational Returns to Multinational Technology Vendors"
Deepa Mani, Ramkumar Janakiraman, Shantanu Dutta and Thomas Dotzel*

Outsourcing or the externalization of business functions to third-party providers has emerged as an imperative for competitive success of firms. Prior research in marketing, strategy and information systems emphasizes the significant incentive and cognitive conflicts inherent to strategic outsourcing and its implications for performance of the outsourcing client. However, there are two sides to a strategic outsourcing contract: the client firm that grants a contract and the vendor firm that receives the contract. Although a rich body of prior work has advanced our understanding of why firms outsource their functions and the effect of outsourcing on their performance, there are few insights on factors that explain heterogeneity in vendor performance. Our study addresses this gap in prior work and examines the characteristics of the outsourcing initiatives that explain heterogeneity in both market and operational performance of vendor firms. Our proposed conceptual framework of vendor performance is built on a set of three factors, namely the type of client (private sector vs. government), contract type (price and length) and vendor experience (client specific, magnitude, and diversity). To empirically test our proposed research questions, we leverage a large and comprehensive dataset of almost 4000 outsourcing contracts implemented by US multinational technology and global vendors in more than 20 countries over a span of 19 years. We find that the market underestimates outsourcing contracts with government clients as well as the benefits of diversity of vendor experience in the host market. Similarly, we find that investors underestimate the advantages of variable price contracts for vendor firms. In terms of the impact outsourcing initiatives have on a vendor’s operational performance, we do not find a systematic difference between government and private sector clients. However, we find significant positive effects of the diversity of vendor experience in the host country and contract length on operational performance. Taken together, the results of the study shed light on a set of contingency factors that can make outsourcing contracts more or less valuable for vendors who seek to work with global clients.

* Deepa Mani is an Assistant Professor of Information Systems and Joint Executive Director of the Srini Raju Centre for IT and the Networked Economy at the Indian School of Business. Ramkumar Janakiraman is an Associate Professor of Marketing and Mays Research Fellow at Texas A&M University. Shantanu Dutta is the Dave and Jeanne Tappan Chaired Professor of Marketing at the University of Southern California. Thomas Dotzel is an Assistant Professor of Marketing at the Desautels Faculty of Management, McGill University. The authors can be contacted via email at Deepa_Mani@isb.edu (Mani), ram@mays.tamu.edu (Janakiraman), sdutta@marhsall.usc.edu (Dutta), and thomas.dotzel@mcgill.ca (Dotzel).

Customer Retention in a Platform World.


Intervenant : Barak LIBAI
Professeur de Marketing , Arison School of Business, Interdisciplinary Center (IDC) – Herzlya, Israel

14 septembre 2017 - Salle T004 - De 10h30 à 12h00

Customer Retention in a Platform World.

By Barak Libai
Arison School of Business, Interdisciplinary Center (IDC) – Herzlya, Israel

In recent year marketer’s attention has been drawn to the prominence of customer retention and the need to low churn, in particular of high value customers. Yet this thinking and analysis has largely focused on the case of a single product, while firms had been advised for a while to manage their offerings as platforms that consist of families of products with a common underlying logic. Taking a platform view, when a new product of the same family enters the market, the users of current products may be good candidates for acquisition efforts to the new offering. This raises interesting questions which widen the existing discussion of customer retention. Which customers would the firm want to transfer from current products and when? Would we want to churn the “best customers” of the current product earlier or later? How would considerations of value that stems from purchases (lifetime value) be different value that stems from the effect on others (social value)? We analyze these questions using agent based models, looking in particular at the market for mobile games, where the question of customer transfer among products is of high importance to firms.

Does the Opinion of the Crowd Predict Success? Evidence from Crowdsourcing


Intervenant : Anirban MUKHERJEE
Professeur Assistant de Marketing , Lee Kong Chian School of Business - Singapore Management University

7 juillet 2017 - Salle T015 - De 10h30 à 12h00

Does the Opinion of the Crowd Predict Success? Evidence from Crowdsourcing


Anirban Mukherjee
Assistant Professor of Marketing
Lee Kong Chian School of Business, Singapore Management University

Ping Xiao
Assistant Professor of Marketing, Visiting Assistant Professor of Marketing
Business School of the National University of Singapore, NYU Shanghai

Li Wang
Assistant Professor of Marketing
Shanghai University of Finance and Economics

Noshir Contractor
Jane S. & William J. White Professor of Behavioral Sciences
McCormick School of Engineering & Applied Science, the School of Communication and the
Kellogg School of Management at Northwestern University


“Crowdsourcing” is the sourcing of organizational functions from the “crowd”: a large, undefined community of a firm’s consumers, partners, and collaborators. A crucial challenge in crowdsourcing is to determine the quality of crowdsourced submissions. To help screen submissions, crowdsourcing portals use crowdvoting: they ask the community to vote on submissions. Our study investigates the informational role of crowdvoting on design submissions on Threadless, a pioneering crowdsourcing website. We collect and examine a novel, large scale dataset tracking over 150,000 designs, submitted by over 45,000 designers, voted on almost 150 million times, by over 600,000 different users. We focus on two questions. First, what is the conventional wisdom—how does crowdvoting influence Threadless? Second, does the conventional wisdom stand up to scrutiny—does crowdvoting systematically predict commercial success? We document several new empirical findings relating crowdvoting to revenues. We conclude by discussing the implications of our research for designers and firms seeking to ride the crowdsourcing tide.

Keywords: crowdsourcing, crowdvoting, new product development, big data.

Technology and Consumer Behavior


Intervenant : Jonathan LEVAV
Professeur Associé de Marketing , Stanford Graduate School of Business

12 mai 2017 - Salle T015 - De 10h30 à 12h00

Technology and Consumer Behavior

By Jonathan Levav

Electronic devices are assumed to make markets more efficient and to create a distribution channel for market information. These devices and the applications that run them allow people to engage in commercial transactions on the go, to access information from all parts of the globe, and to communicate through voice and video from anywhere. In the series of papers that make up this talk I will show that people's interactions with these devices can evoke psychological processes that influence the judgments and decisions that people make when using them. Specifically, in multiple field and lab studies I examine how the physical interaction with electronic devices influences psychological processes in systematic ways.

Online Streaming and its Effects on Society


Intervenant : Hannes DATTA
Professeur Assistant - Département de Marketing , Tilburg School of Economics and Business

21 avril 2017 - Salle T015 - De 10h30 à 12h00

Online Streaming and its Effects on Society

By Hannes DATTA
Joint work with George Knox and Bart Bronnenberg (both Tilburg University)

Digital streaming is set to take over as the dominant business model in industries like music (e.g., Spotify), movies (e.g., Netflix), books (e.g., Kindle Unlimited), and games (e.g., Steam). Instead of purchasing individual content, streaming allows users to rent access to a vast library of digital content that is free at the margin. Using a panel data set of individual consumers’ listening histories across many platforms, we study how the shift from purchasing to streaming affects society.

Prior work has established that the adoption of online streaming leads to a sizeable effects at the individual level. For example, consumers discover more new content, and tend to favor less popular artists over superstars. However, it is not clear how online streaming affects consumption behavior at the societal level (i.e., across consumers). On the one hand, consumer tastes may become fragmented when choosing among less popular and newer artists. On the other hand, consumer tastes may become more homogenous if recommendation systems and curated playlists on streaming services push users to the same new content.

From a public policy perspective, too much fragmentation is bad news because it can diminish social capital, as fewer people share the same experience. However, too little fragmentation may signal a lack of diversity, favoring superstars and damaging independent labels. We examine several possible drivers of fragmentation, and use our data to test competing explanations.

Valuing Non-Contractual Firms Using Common Customer Metrics


Intervenant : Daniel McCarthy
Doctorant Départment de Statistiques , Wharton School of the University of Pennsylvania

3 mars 2017 - Salle T04 - De 10h30 à 12h00

Valuing Non-Contractual Firms Using Common Customer Metrics

There is growing interest in the notion of “customer-based corporate valuation,” explicitly tying the value of a firm's customers to the firm's overall financial valuation. While much progress has been made in building a well-validated customer-based valuation model for contractual (or subscription-based) firms, there has been less progress for non-contractual firms (e.g., retail, travel/hospitality, and mobile gaming). Non-contractual businesses have more complex transactional patterns than contractual ones for a variety of reasons, including (1) they are characterized by latent attrition instead of observable churn behavior, (2) they often have irregular purchase incidence timing and spend amounts. These factors make it harder to reconstruct granular purchase behaviors from aggregate data, and to understand what metrics would serve as the best inputs for such a model. Despite this lack of guidance, a number of non-contractual firms regularly report a variety of different aggregate measures to their shareholders (e.g., the number of active users). We use a novel methodology based upon “indirect inference,” a well-established generalization of generalized method-of-moments procedures, to draw a connection between these common aggregate metrics and the underlying parameters of latent variable models for repeat purchasing. We show how the overall predictive validity of the models varies as a function of the combination of metrics used to train the models; this allows us to better understand both how many and which metrics are needed to achieve adequate predictions of future revenues. We apply this methodology to quarterly data from the largest subsidiary of an e-commerce retailer, valuing the subsidiary as a whole, decomposing this valuation into existing and yet-to-be-acquired customers, and analyzing the profitability of newly-acquired customers.