AI-Driven Customer Persona Discovery in Retail


SoMin worked with one of the World's Largest "Mother & Baby" Retail brand represented in Indonesia. The Client was looking to determine the most effective (in terms of both click-through-rate (CTR) and cost-per-click (CPC)) Facebook targeting options for their Hari Raya e-commerce sales campaign.


By analyzing over 30 million data points discovered from 148 data sources, has identified 32 highly relevant AI-Driven Facebook Audiences for targeting the same Ad Copy over a 6-day Facebook Split Test campaign. The Ad was shown in 11 different placements and jointly optimized by CPC, CTR, and Impression delivery pace.

The test results have been benchmarked against the client's previous campaign results as well as the traditionally-set targeting (demographic-based targeting, lookalike targeting, page audience-based targeting, competitor-based targeting). 


Based on the test results, was able to discover AI-Driven Facebook Targeting Audiences that are 40% more relevant (in terms of CPC) and deliver results at 25% lover cost (in terms of CPC).

The campaign results have also demonstrated that AI Audiences demonstrate:

  • 50% average higher CTR vs traditional targeted audiences.
  • 40% average lower CPC vs traditional targeted audience.

We’d like to set cookies to help us improve our site by collecting and reporting information on how you use it. These cookies collect information that does not directly identify anyone. For more information on how these cookies work please see our Cookies Policy.