Every media company today is under increasing pressure to optimize their marketing dollars to retain and increase viewership. Discovery, Inc. is a global leader in entertainment, with 19 channels available to millions of US households including HGTV, TLC, Food Network, Investigation Discovery, Oprah Winfrey Network and more. Discovery needed a data-driven solution to allocate hundreds of millions of dollars across its media platforms in the United States.
With media planning responsibilities spread across various networks and brand initiatives, Discovery wanted to pair the expertise of their 20 media planners with a machine learning solution that could provide a scientific recommendation on how to allocate spend across television, social, digital, search, radio, print and out-of-home placement opportunities in order to maximize ROI.
Civis partnered with Discovery’s team to design and build a custom web application called “Athena,” which empowered the organization’s strategists to more efficiently and rigorously make data-informed decisions.
Athena provided immediate visibility into how to optimize media allocation and budget based on show characteristics and historical marketing performance. Most importantly, it allowed media planners to simulate dozens of media allocation scenarios for one campaign and choose the one that maximized return on investment.
Because Athena was built on top of the Civis Platform, it utilizes machine learning models to consider interactions across a premiere’s characteristics and media spend on different channels, and it enables easy team collaboration and robust security.
Since Discovery adopted Civis’s Athena tool, Discovery has cut their cost of acquisition in half. More than 20 planners and analysts now use the tool multiple times a week for 100+ campaigns each year.
We can elevate the decision making at the company around optimization, tradeoffs, and scenario planning. The tool has made us orders of magnitude more advanced … a total triumph.”
–Seth Goren, SVP of Media Strategy & Analytics at Discovery