On April 17th, the Volcano Engine Data Flywheel Industry Seminar was held in Beijing. This conference focused on two major data scenarios in the internet industry: data asset construction and business data application. Drawing on ByteDance's over a decade of experience in data-driven practices, the seminar explored the application scenarios and enterprise practices of the new model of digitalization upgrade, the "Data Flywheel," in the internet industry.
The head of Volcano Engine Data Product Solutions introduced that the Data Flywheel, as one of the mainstream models for current domestic enterprise digitalization upgrade, has been practiced in multiple industries including internet, finance, automotive, and retail. Internet enterprises, benefiting from the natural digital environment, can better adapt the Data Flywheel to enhance "systematic and comprehensive data asset construction, and accurate and efficient business data application."
According to the 53rd Statistical Report on Internet Development in China recently released by the China Internet Network Information Center (CNNIC), as of December 2023, the number of Chinese internet users reached 1.092 billion, with an increase of 24.8 million compared to December 2022, and an internet penetration rate of 77.5%. For internet enterprises, this represents both a new opportunity for business growth and a new challenge based on data-driven insights, business decisions, and full-service delivery.
Facing these new opportunities, what kind of data do enterprises need? "First and foremost, it must revolve around what the business needs," mentioned the architect of Volcano Engine Data Product Solutions during the sharing session. Building data assets based on "business needs" is the foundation for a company to ultimately implement data-driven practices. This is also a concrete expression of the Volcano Engine Data Flywheel's promotion of richer data assets, optimized data quality, and more efficient data development starting from "data consumption."
Taking ByteHouse, the cloud-native data warehouse product independently developed by Volcano Engine, as an example, it can meet the actual business needs of enterprises by realizing containerization, storage-compute separation, multi-tenant management, and read-write separation. It possesses advantages over similar products in the market in terms of scalability, stability, operability, performance, and resource utilization efficiency.
Once the enterprise's data assets are established, DataFlywheel provides its own model for helping businesses leverage data more effectively and efficiently. Centered around "data consumption," DataFlywheel offers intelligent data insights such as DataWind for smart insights, DataTester for A/B testing, and DataFinder for growth analysis. These data intelligence products assist employees in making more informed business decisions, improving operational efficiency, and ultimately enhancing business value.
An O2O real estate platform in China utilized the series of products offered by the DataFlywheel model through its Volcano Engine. Precise insights into customer needs, categorized demand tags, and tailored push notifications designed for different customer needs have effectively boosted customer engagement and conversion rates on the platform.
During the event, Wu Xianbin, the project leader for real-time design, shared the latest practices of DataFlywheel in user retention scenarios within the real-time design community. "Community-based operations are the highlight of real-time design," he explained. "Users can freely browse and search for various design materials such as works, components, icons, and illustrations within the community."
Through data insights, real-time design discovered that users who have utilized design materials are more likely to stay within the community, while another set of data highlighted the importance of the search scenario. "This provided us with some insights," Wu Xianbin remarked.
With the idea of iterating and upgrading existing search strategies to potentially increase user retention and engagement within the community, Wu Xianbin and his team utilized DataTester to launch A/B experiments, ultimately resulting in significant optimization effects.
From ByteDance's own experiences to industry-specific solutions focusing on the internet sector, and now the latest practices in real-time design, over 30 participating companies experienced firsthand the new insights and changes brought by DataFlywheel to industry-wide data intelligence upgrades. Today, the Volcano Engine has widely applied this new model across multiple industries, empowering enterprises to achieve growth.
