Motivating user-generated content with performance feedback: Evidence from randomized field experiments

Ni Huang, Gordon Burtch, Bin Gu, Yili Hong, Chen Liang, Kanliang Wang, Dongpu Fu, Bo Yang

Research output: Contribution to journalArticlepeer-review

71 Scopus citations


We design a series of online performance feedback interventions that aim to motivate the production of user-generated content (UGC). Drawing on social value orientation (SVO) theory, we develop a novel set of alternative feedback message framings, aligned with cooperation (e.g., your content benefited others), individualism (e.g., your content was of high quality), and competition (e.g., your content was better than others). We hypothesize howgender (a proxy for SVO) moderates response to each framing, andwe report on two randomized experiments, one in partnership with a mobile-app-based recipe crowdsourcing platform, and a follow-up experiment on Amazon Mechanical Turk involving an ideation task.We find evidence that cooperatively framed feedback is most effective for motivating female subjects, whereas competitively framed feedback is most effective at motivating male subjects. Ourwork contributes to the literatures on performance feedback and UGC production by introducing cooperative performance feedback as a theoretically motivated, novel intervention that speaks directly to users' altruistic intent in a variety of task settings. Our work also contributes to the message-framing literature in considering competition as a novel addition to the altruism-egoism dichotomy oft explored in public good settings.

Original languageEnglish (US)
Pages (from-to)327-345
Number of pages19
JournalManagement Science
Issue number1
StatePublished - Jan 2019


  • Gender
  • Performance feedback
  • Randomized field experiment
  • Social value orientation theory
  • User-generated content


Dive into the research topics of 'Motivating user-generated content with performance feedback: Evidence from randomized field experiments'. Together they form a unique fingerprint.

Cite this