In a world where e-commerce platforms offer consumers an overwhelming array of products, data assets and artificial intelligence algorithms are the linchpins of digital platform operations as data-driven algorithm recommendations help users discover and purchase their desired items. But what would happen if these recommendations were turned off?
Professor Sun Tianshu, CKGSB Dean’s Distinguished Chair Professor of Information Systems and Director of the Center for Digital Transformation, has explored this question by quantifying the value of personal data to e-commerce platforms through the world’s first large-scale randomized field experiment on 550,000 e-commerce users in collaboration with Zhejiang University and Alibaba, detailed in his recent paper published at Management Science, a top management journal, titled, “The Value of Personal Data in Internet Commerce: A High-Stake Field Experiment on Data Regulation Policy.”
The experiment uniquely assessed the significance of data assets to the digital platform economy by disabling the algorithmic recommendation system based on personal data on the platform’s homepage for the test group. The results were telling: when users were denied access to information flows recommended to them based on characteristics inferred from their personal data, ecommerce platforms saw asymmetry in the products recommended, hurting the diversity of products on the platforms and leading to fewer transactions and lower engagement rate.
Products recommended on the homepage dropped from 4 million to around 280,000, and the top 1,000 products from leading brands took up 90% of recommended slots for a disproportionately high exposure. Customers’ click-through rate (CTR) and page views (PV) went down by 75.3% and 33.6% respectively and sellers’ gross merchandise value (GMV) and transaction volume plummeted by a whopping 81.1% and 86%. Small and medium-sized sellers on the platform saw a 54% decrease in pageviews and a 90% drop in GMV, suggesting they were more impacted than leading sellers, whose PV fell by 27% and GMV by 79%. Research also found consumers with low purchasing power are negatively impacted.
This paper provides a quantified basis for how to balance the use of data assets and personal privacy protection through large-scale experiments, but also helps businesses formulate more informed digital economic policies.
Professor Sun leads CKGSB’s Center for Digital Transformation, focusing on digital business, digital technology and implication for traditional businesses. This research underscores CKGSB’s commitment to pioneering knowledge that drive understanding and innovation in the digital economy.
