The precision marketing system of KAOLA.COM allows merchants to batch select customers, configure marketing strategies, and view marketing effects. By using the algorithm capabilities, merchants can tailor their marketing campaigns to their customers and recommend products to potential customers based on scenarios. The customer information system can ingest both batch data and streaming data. Batch data is synchronized from MaxCompute to AnalyticDB for MySQL on a regular basis. Then, the synchronized data and the primitive data stored in AnalyticDB for MySQL is used to generate data packages that contain information about customers for business systems of KAOLA.COM. All similar subqueries within the system can be accelerated by using materialized views.
Materialized views can be created based on high-frequency subqueries, and the ON OVERWRITE refresh mechanism is used to update the materialized views at a specific interval. When data is synchronized from MaxCompute and overwrites data in AnalyticDB for MySQL tables, relevant materialized views are automatically updated.
The automatic rewriting feature can also be enabled on materialized views. This converts SQL query statements into queries of materialized views, which significantly reduces computing time. The time required for a query operation can be shortened from 20 seconds to less than 500 milliseconds.