Although traditional digital media strategies had been effective we knew using the right data could make us even more efficient and improve performance. • Working directly with US-based Chinese language sites provided accuracy of the Chinese American audience. However, planning/buying 20+ sites to achieve any type of scale proved to be inefficient. • Programmatic buying was our best option to increase overall reach but instead of relying solely on 3rd party data to reach the Chinese segment, we gathered and created our own audience data to complement what was available through 3rd parties. We were able to identify Chinese Americans and deliver to them the right in-language message at the right time and to the most people.

• Step 1: Utilizing a data management platform (DMP) and a self-service demand side platform (DSP) we worked with a set of preferred partners to create our own 1st party Chinese-speaking audience and targeted them with the State Farm message.

• Step 2: From there we were able to identify specific behaviors common to those showing the most interest in purchasing auto and/or home insurance.

• The Closer: Using this information, we expanded our audience and reach by combining our 1st party data with 3rd party data and targeted similar looking audiences.

• Armed with additional data that improved accuracy, context and scale, we launched our revamped programmatic campaign from 1/10/16 - 31/12/16.

• We nearly doubled the number of Chinese Americans likely in-market for auto or home Insurance that we were able to reach through our campaign. • As a result we also saw a 340% increase in conversion rate among those reached through the campaign (0.22%), compared to non-data targeted partners (0.05%).