Driving Better C2C Seller Performance for One of the Largest eCommerce Marketplaces in the World Prescience Decision Solutions January 6, 2023

Driving Better C2C Seller Performance for One of the Largest eCommerce Marketplaces in the World

The customer operates one of the largest eCommerce marketplaces in the world with both Business-to-Customer (B2C) and Customer-to-Customer (C2C) shopping / auction portals. One of the key challenges was a lack of specific insights to track and improve C2C performance.


The objective of the initiative was to drive better performance of C2C sellers through better understanding of the different personas and identification of key behavioral aspects that can be used improve outcomes for the client.

Another major challenge was to retain high-performing sellers who were familiar with the platform. A delisting by such a seller had a 3x negative impact on sales as compared to sellers who were unfamiliar with the platform.



Using the data collected from C2C marketplaces, the Prescience team was able to build detailed profiles of each persona and their corresponding selling behavior. In addition to this, we also identified certain key areas where seller performance could be improved. This initiative helped the company achieve two key goals:

  1. Improve GMV (performance of sellers)
  2. Decrease churn and address deceleration in seller activity
We used data analytics, qualitative research, and quantitative research to answer key questions about the sellers and their needs. Our approach to solving the problem helped deliver positive business impact:
  • We performed hypothesis testing to understand if familiarity with the platform would positively impact C2C sellers’ success. The team also wanted to determine if familiarity is determined by both selling and buying aspects together.
  • Familiarity and selling behviour data was used to segment sellers into 7 clusters. Our team considered the selling and browsing KPIs to define the thresholds of the clusters while conducting the familiarity assessment.
  • A cluster analysis was performed across stores, verticals, life-stage segmentation, and completion rates of listings and gave us prescient insights.
    • Provided personalized recommendations for each cluster group based on various parameters for better overall engagement. Developed deep-dive recommendations for specific clusters, such as focusing on engagement for familiar users and guided navigation through the platform for those who are not familiar with the platform, to avoid high churn rates.
    • After analyzing the cluster movements and churn risk involved, we suggested intervention techniques to reduce the churn rate
  • We formulated a modified technique of engaging more C2C sellers with our platform (handholding, incentives on selling through buying, etc.) that resulted in better engagement from sellers in the C2C segment. We also identified clusters that could offer significant additional business opportunities.
  • A set of recommendations was arrived at in order to decrease churn and deceleration among seller clusters and other segmentations. As a result, 25% of the earlier high-risk sellers with a lower familiarity with the platform were migrated to the familiar/high GMV stage with limited churn risk.
  • Sellers with higher familiarity with the platform were also seen to be one of the highest performers. We helped the company improve the selling experience through tailored recommendations for these high-performing sellers.

Overall, our efforts helped the company improve its performance in terms of GMV and decrease the churn and deceleration rates among its sellers. The team’s data-driven approach and personalized recommendations led to better engagement from both sellers and buyers, which in turn helped improve the company’s bottom line.

Learn from the Success Stories of Business Transformations Fueled by Data, AI and Machine Learning