Based on the logistic regression algorithm and the interests of the seed population, a similar population can be predicted according to the model. This builds groups of similar customers based on as many as 3 billion data points.

Our team of professional scientists has built over 800 tags in 6 categories based on 250 million daily data points, which allows for combining tags to build target groups.

Geofences are set up to provide information on business districts, proximity to hospitals, and proximity to transportation hubs in a specific area. Then this information is used to build target groups.

A unified ID supports target group matching between 6 types of service IDs, which protects data privacy while quickly building target audience segments.