MediaWave Actionable Insights and Industry News for Media Professionals
Two toy robots marching down the street

Humans with (not vs.) Machines

July 18th, 2019   ||    by Christy Rosell

Media planners (and their clients) win with ad tech

A news piece called “Manufacturing Music” on the “Today Show” explored where artificial intelligence (AI) is taking the music industry. It featured an interview with Lucas Cantor, a Los Angeles-based composer, who used AI to complete Franz Schubert’s unfinished “Symphony No. 8” from 200 years ago. Classical music fans have been baffled by how it should have ended. With AI, the composer analyzed Schubert’s other symphonies for writing patterns to more accurately predict how the completed symphony may have ended. This was exciting for many and caused anxiety in others, who feared the end of human-composed music.

But Cantor doesn’t think machines will take over music composition. He looks at it differently. He said AI makes him stronger, with access to a collaborator who doesn’t get tired and has an endless stream of ideas.

“The purpose of AI is to offload some of the analytical tasks that computers are better at than human minds,” he said.

The focus at Videa is similar. Rather than AI, Videa harnesses the power of machine learning to help media planners forecast TV ratings and automate smart ad buys. Watching this news piece got me thinking about technology in the ad industry and asking myself these questions:

  • Has the pace of change sped up?
  • Are things better now?
  • Who benefits?

(Answers to some of these questions and more insights can also be found in Videa’s “Join the Movement” study of agency and marketing professionals.)

And like Lucas Cantor, the general outlook is pretty optimistic.

Machine learning & AI explained

Despite recent media attention, the concept of artificial intelligence (AI) has been around since 1936, when scientist Alan Turing invented his “Turing machine,” essentially a calculator. It was the machine that enabled the Allies during WWII to break the Nazi’s secret codes used to encrypt their communications. Ultimately, his work laid the foundation for the field that would later become known as computer science, and the concept of AI was there from the beginning.

We generally think about machine learning as a subset of AI. We also recognize the difference in the history of the development of these disciplines. As mentioned earlier, AI comes from theoretical computer science while machine learning really grew out of statistics. Many techniques employed today combine concepts from each of these disciplines’ historical origins.

In recent years, our society has started to leverage AI and machine learning in an increasingly large number of areas and at a faster pace. For example, social media companies are leveraging such techniques to perform facial recognition in images as well as recognizing and reporting hate speech. Additionally, brands are creating virtual or augmented reality experiences for customers. The evolution of cloud technology has only catalyzed employing these techniques more broadly.

Ad tech for media planners and buyers

The growth of AI and machine learning has also impacted advertising technology. A recent example is AI and machine learning-enabled bidders that fairly and efficiently place orders in the digital space. As has traditionally happened in the past with technological adoption, the digital space has been an early adopter.

But we feel the time has now come when these technologies can enable broadcast TV processes. Videa, for example, focuses on how to use AI and machine learning techniques to streamline the local TV media buying process. And there have been some compelling results.

Most recently, Videa focused on one of the biggest pain points for media buyers: ratings forecasts. (Click here to download their ratings study.) We know buyers spend about 25 percent of their time doing the math to manually forecast ratings. Videa has figured out how to improve buyers’ efficiency without sacrificing accuracy by employing machine learning approaches, using data from Nielsen and Comscore across 210 DMAs (designated marketing areas). Recently, they compared their own forecasts to agency buyers’ forecasts that were calculated the traditional way. Videa was not only able to deliver all of the forecasts in seconds, but when both forecasts were compared to the actual ratings delivery, Videa’s forecasts were slightly more accurate in aggregate.

Some agency personnel hear this and fear losing their jobs to technology – and Videa’s “Join the Movement” study reflected some of this trend.

But with machine learning, more tasks can be accomplished. Look at digital buyers. They didn’t exist 20 years ago. In this space, technology continues to improve and find new ways to automate their work and improve buying experience for this pool of inventory. The net effect is higher demand for people with these skill sets. Ad tech, such as the automated bidders, allows digital buyers to spend more time on value-added activities.

Soon, broadcast TV buyers and planners will spend less time doing simple and monotonous calculations and instead focus on higher value activities such as analyzing campaign posts, complying with campaign audits, ensuring rotation fairness and developing a composite for the broader campaign, because they finally have the time.

It can be the same for anyone in the TV and advertising industries. And instead of feeling confronted by the next great ad tech, we can all feel stronger, like Lucas Cantor. Because choosing the right tech-enabled teammates or collaborators means we can all focus on activities with deeper meaning that truly help advertisers win.

 

 

Tags: , , ,
arrow_upward