Using Data and AI to inform creative: Josh Siegel

By Christi Warren

In his role at Gannett | USA TODAY NETWORK, Josh Siegel leverages audience data and AI to better inform advertising design, to answer the question: What makes an advertisement good?

As general manager and senior vice president of the subscription and data products group, he leads his team to understand the answer to that question through a simple metric: click-through rate. Siegel spoke at the 10^th Data & AI for Media Week at Microsoft in San Francisco on May 3 about driving click-through rates with AI-optimized advertising creative.

But it’s not nearly as simple as it sounds. To understand that rate, Siegel and his team used computer vision and deep learning techniques to analyze over 50,000 ads to see what factors might drive improvements.

Through that analysis they developed insights and a model to determine an ad’s effectiveness across four factors:

• Objects/style used in creatives
• Colors used in the ads
• Text, including the amount and placement
• Words used in the “call to action”

Some takeaways were: simple and direct ads that aren’t too cluttered are helpful, and if you’re going to use animation make it clean and to the point. If you’re a print organization, and you’re trying to turn a print ad into a digital ad, it’s not going to work.