Mediahuis uses machine learning and brain patterns to predict customer behaviour

By Shelley Seale

INMA

Austin, Texas, United States

Connect      

Jessica Bulthe of Mediahuis knows quite a bit about how the brain works. As a student of psychology, she has studied the human brain with some interesting results, particularly when it comes to numbers.

She shared some of this during INMA Media Innovation Week in Hamburg, Germany.

“One of the most interesting things in how the brain works is knowing, for example, if you see the number two in Arabic [numeral] or you see two apples — does this mean the same number to the brain or not?”

To answer this question, Bulthe turned to Artificial Intelligence for her study.

“We simulated the brain applied models,” she said. One of the most relevant findings that she applies to her job with Medihuis is that machine learning algorithms can predict brain patterns based on certain numbers that the subject has seen.

“Now we predict when somebody’s going to churn, and when somebody might be interested in buying a certain product from advertisers,” she said. “It’s the same kind of analysis, but in a different way.”

Bulthe has seen some changes in how Medihuis works with data during her tenure there.

“When I started with the company, we just had a DMP for only a month, so we didn’t do anything with all the data that we had, for advertisements or for anything else.”

In the last two years, however, the organisation has created a much more robust data analysis model — one that looks at specific audience segments such as sports fans and car lovers.

“Now we’re drawing really deep level custom work for our advertisers,” she said. “We place pixels, we match databases, or we link DMPs to actually create custom audiences for our advertisers. In a year and a half, we switched from no DMP whatsoever to really cool stuff in advertising.”

About Shelley Seale

By continuing to browse or by clicking ‘I ACCEPT,’ you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.
x

I ACCEPT