Data-Driven Entertainment: iQIYI’s Big Data Platform and AI Applications

Netflix’s Todd Yellin once described the company’s success as a double helix that tightly intertwines technology and creative entertainment. With over 100 million members and more than 14 billion hours of daily viewing, Netflix demonstrates the massive potential of data in video entertainment.

In China, iQIYI stands out as a leading video entertainment company that has leveraged data to fuel its rapid growth. After launching hit shows such as “The Rap of China” and “Idol Producer,” iQIYI went public in March and reported 67.1 million members in Q2, a 75 % year‑over‑year increase.

The key question is: what role does data play in this success?

iQIYI’s data team, led by Sun Bin, uses data analysis to identify market potential, such as recognizing the latent demand for hip‑hop in China before launching “The Rap of China.” By analyzing on‑site behavior (views, comments, danmu) and off‑site public sentiment with natural language processing and big‑data techniques, they confirmed a strong, youthful hip‑hop fan base.

To avoid data silos, iQIYI built a unified big‑data platform nicknamed “Tongtian Tower.” This full‑stack organization integrates data collection, processing, analysis, and application within a single centralized team, enabling both internal business intelligence and external personalized recommendation services.

The platform provides three core data capabilities: (1) intuitive visualizations (numbers, charts); (2) self‑service analytics tools with drag‑and‑drop interfaces and custom SQL; and (3) an open API for developers. Product managers and analysts are required to understand both business and data to ensure that data solutions meet content‑team needs.

Two concrete use cases illustrate the platform’s impact:

1. Smart Casting System – By creating a standardized video résumé database and applying AI for natural‑language and visual understanding, the system matches script roles with actor profiles, dramatically improving casting efficiency. Human directors provide feedback to refine AI selections, creating a loop of supervised learning.

2. Paid‑Member Churn Prediction – The data team builds behavioral models to flag members at risk of leaving. Early warnings allow the operations team to intervene, contributing to a Q2 revenue of ¥2.5 billion and a 66 % YoY increase in subscription revenue.

iQIYI’s leadership, including CEO Gong Yu, emphasizes a data‑driven culture. Product and technology staff must constitute at least half of the workforce, and data‑centric decision‑making is embedded in regular business reviews.

Data scientists at iQIYI are expected to understand specific scenarios (e.g., weekly vs. daily releases, view‑count vs. trend metrics) and to combine data insights with domain expertise to guide content investment and production decisions.