Yupeng Chen

Yupeng Chen
  • Doctoral Candidate in Marketing

Contact Information

  • office Address:

    700 Jon M. Hunstman Hall
    3730 Walnut Street
    University of Pennsylvania
    Philadelphia, PA 19104

Research Interests: Referral Programs, Preference Estimation

Links: CV


Yupeng is a rising fifth-year doctoral candidate in Marketing at the Wharton School of the University of Pennsylvania. His research interests lie in two areas: referral programs and preference estimation. He is particularly interested in conducting field experiments to identify strategies that can enhance the effectiveness of referral programs and understand the underlying mechanisms. He is also interested in developing machine learning methodologies for accurate estimation of consumers’ heterogeneous preferences. His research has been published in Marketing Science.

Prior to Wharton, Yupeng obtained his Ph.D. in Operations Research from Columbia University in 2015 and his B.S. in Mathematics from Peking University in 2009.


Continue Reading


  • Yupeng Chen (Working), Enhancing Effectiveness of Referral Programs by Promoting Better Matching: Evidence from Field Experiments.

  • Yupeng Chen (Work In Progress), A Low-Dimension Learning Approach to Modeling Consumer Heterogeneity in Choice-Based Conjoint Estimation.

    Abstract: Estimating consumers’ heterogeneous preferences using choice-based conjoint data is challenging since the amount of information elicited from each consumer is often limited. Consequently, effective modeling of consumer heterogeneity becomes critical for accurate conjoint estimation. We propose a low-dimension learning approach to estimating consumers’ heterogeneous preferences and apply it to choice-based conjoint estimation. The intuition behind the proposed approach is that, by restricting the individual-level preference vectors to a low-dimensional linear manifold, we are able to focus on a small number of important orthogonal directions of preference variations and effectively utilize choice data to recover preference variations along such directions. We develop a convex optimization formulation to operationalize this intuition that builds on recent advances in rank minimization and machine learning. We evaluate the empirical performance of the proposed low-dimension learning approach using both simulation experiments and field choice-based conjoint data sets.

  • Yupeng Chen and Meng Li (Work In Progress), Probabilistic Referral Rewards: Do They Work and Why?.

    Abstract: Many firms face the challenge to their referral programs that, while modest rewards are not effective in incentivizing referrals, their marketing budgets constrain them from increasing the rewards. We propose probabilistic referral rewards as a potential solution to address this challenge. While standard economic models assuming risk neutrality or risk aversion predict that probabilistic referral rewards cannot be more effective than deterministic referral rewards of the same expected value, a pilot study conducted at a Chinese e-commerce platform suggests that the former could be more effective than the latter. We have secured the cooperation of the e-commerce platform to conduct field experiments of a larger scale to assess the effectiveness of probabilistic referral rewards and explore the underlying mechanisms (e.g., risk-seeking and optimism).

  • Yupeng Chen (Work In Progress), Value-Based Referral Rewards, Motivation Crowding-out, and Opportunistic Referrals.

    Abstract: In this project, we study the strategy of rewarding referring customers based on the value of their referred customers. In particular, we investigate the impact of the design of value-based rewards on customer referral behavior and the effectiveness of referral programs. We are particularly interested in understanding whether and when value-based rewards would crowd out customers’ intrinsic motivation for referring friends, as well as when and which customers would exploit value-based rewards by making opportunistic referrals. We have secured the cooperation of a Chinese online financial services firm to conduct field experiments to answer these questions.

  • Yupeng Chen, Raghuram Iyengar, Garud Iyengar (2017), Modeling Multimodal Continuous Heterogeneity in Conjoint Analysis - A Sparse Learning Approach, Marketing Science, 36 (1), pp. 140-156.


Latest Research

Yupeng Chen (Working), Enhancing Effectiveness of Referral Programs by Promoting Better Matching: Evidence from Field Experiments.
All Research