Andreea Popescu, Director, Statistics and Predictive Analytics
Turner Broadcasting System
For decades, Turner Broadcasting used viewer demographics such as age and gender to target audience segments in order to align viewers with appropriate TV ad spots. The approach, however, did not maximize potential ROI for the ads, so Turner wanted to develop an in-depth understanding of client needs, behaviors and objectives in order to re-align ad spots with an optimal viewer demographic.
Turner developed a spot scheduler application consisting of a suite of advanced analytics tools. The application’s descriptive component uses analysis and visualization tools to present detailed historical or predictive information for particular segments and allows comparison of Turner networks against competitors. Its predictive component includes a scalable, accurate and data-agnostic forecasting approach called Competitive Audience Estimation, which can build granular estimates for any audience segment. The prescriptive component integrates large-scale, mixed-integer-programming models to optimize deal composition and spot placements. Turner matches output from the spot scheduler application with customer segments created from credit card and survey information in order to determine where to place commercial spots across networks.
To date, the solution has generated more than 1,000 ad deals and has increased ad revenues more than 450% per year. Through the development of the spot scheduler application, Turner has created a data-driven culture that will foster alignment and support for analytics initiatives across the organization.