Skip to Content Skip to Footer
Platform Exploitation: When Service Agents Defect with Customers from Online Service Platforms

Platform Exploitation: When Service Agents Defect with Customers from Online Service Platforms

JM Insights in the Classroom

Teaching Insight:

This research investigates the problem of platform exploitation faced by service platforms, in which service agents knowingly break platform rules to entice customers off platform and transact with them directly. It shows service agents who are high quality or with long platform tenure (i.e., they’ve been using the platform longer) are more likely to entice customers off platform, and these agents also return to the platform more slowly to fill future orders, since they spend more time taking care of their off-platform customers. It also tests financial and social strategies that can reduce platform exploitation.

Access Classroom Lecture Slides

Related Marketing Courses: ​
Digital Marketing; Marketing Strategy​​​​;

Full Citation: ​
Zhou, Qiang, B.J. Allen, Richard T. Gretz, Mark B. Houston (2021), “Platform Exploitation: When Service Agents Defect with Customers from Online Service Platforms,” Journal of Marketing.

Article Abstract
The authors conduct a meta-analysis on the effect of electronic word of mouth on sales by examining 51 studies (involving 339 volume and 271 valence elasticities) and primary data collected on product characteristics (durability, trialability, and usage condition), industry characteristics (industry growth and competition), and platform characteristics (expertise and trustworthiness). Their analysis reveals that electronic word-of-mouth volume (valence) elasticity is .236 (.417). More importantly, the findings show that volume and valence elasticities are higher for privately consumed, low-trialability products that operate in less competitive industries and whose reviews are carried on independent review sites. Furthermore, volume elasticities are higher for durable goods and for reviews on specialized review sites, whereas valence elasticities are greater for community-based sites. Drawing on the results, they discuss several implications for managers and researchers and explain why valence elasticities are often found to be insignificant. Finally, they propose numerous directions for future research in the area on the basis of their findings.

Special thanks to Holly Howe (Ph.D. candidate at Duke University) and Adam Mills (Assistant Professor of Marketing, Loyola University), for their support in working with authors on submissions to this program. 

Search other Insights in the Classroom​

Read a managerial summary of this paper.

More from the Journal of Marketing​​​​​​​