How AI could make A/B testing a thing of the past

Traditionally, building effective ad creative has involved a lot of human guesswork, but AI technology is on the cutting edge of offering audiences ads they’ll actually enjoy.

Date published
August 10, 2018 Categories

Even though one of the main goals of digital marketing is to serve customers the right message at the right time, we all know what it’s like to be chased around the internet by an ad that’s completely irrelevant or just plain annoying. And that’s because deciding when and where to deliver that message, not to mention the labor that goes into creating the message in the first place, has long involved human guesswork. Granted, those guesses often come after rigorous testing, but those best tests are limited to the often slow-moving process human analysis. Until now.

Artificial intelligence has already transformed everything from the IT department to the customer service experience, and now, machine learning is on track to completely change the ways we think about ad creative.

Customers only click ads they want to see

A recent study by Adlucent found that 46% of consumers say that their ideal online experience would involve free access to websites that served only relevant ads, with 58% reporting that personalized content improves their perception of a brand.

But until now, the kind of testing that would allow brands to truly deliver ad creative tailored to customers’ preferences has involved heavy reliance on past behaviors or the preferences of similar customers.

However, new advances in AI technology are changing both the ways we test and serve ads. For example, Google recently released an AI tool that automatically adjusts ads to users’ searches, saving advertisers the time they previously spent painstakingly optimizing ads.

And Tel Aviv-based company called Bidalgo has recently released Creative AI, a tool that makes it much easier to inform advertisers about customers’ ad preferences. Soon, AI could transform the way advertisers create and test campaigns.

“Machine learning is going deeper and deeper into the media buying and online advertising process,” says Ran Milo, Vice President of marketing for Bigalgo. “But creative is still the biggest factor in media buying success. However, it’s also the area where advertisers have very little insight and very little data. The solution is to identify which of your creative is actually performing.”

AI can predict what audiences want

Bidalgo’s tool actually scores elements of an ad using KPIs from all parts of the buyer’s’ journey, from the top of the funnel down, and then compares creative performance for different messaging and images, a task that would be virtually impossible without AI.

“Right now, advertisers are flying blind,” Milo says. “Even if you understand that certain creative works better, it’s hard to understand why. We’re using machine learning for image and video recognition to break down different variables, such as concept and copy to find out what’s affecting different KPIs.”

For example, without A, advertisers would need to A/B test by changing just one variable, such as tweaking a headline and running variations against one another before declaring a winner and moving on to testing some other aspect. But AI can test dozens of things at once, which means faster, more accurate data that can serve to make better ads in the future. So if customers are more likely to click on an advertisement featuring mountains and the word “beautiful,” future messaging can build on that feedback.

According to Milo, AI analysis can also inform advertisers as to whether their issue is with quality or quantity.

“We can see if a brand’s creative is really working, and they simply need to invest in producing more content and advertising,” Milo says.

Why do we like what we like? AI might have some answers

Though many companies are adopting machine learning technology, industry-wide data is still pretty thin on the ground, so there’s not much consensus about why customers like what they like. But that information might be coming soon.

“Right now, advertisers can get a lot of insight about what works for them,” Milo says. “As soon as we have more data, we can compare industry-wide trends and really find commonalities for what’s working and what’s not. AI isn’t trying to understand the psychology behind ads yet, but maybe soon.”

And while humans are still primarily responsible for conceptualizing and producing ad content, even that could change in the future. Last year, Shun Matsuzaka of McCann Japan introduced the world to its first robot creative director. He and his team fed AI a database of award-winning advertisements with the goal of creating a commercial for breath mints. The result was a  surreal ad featuring a flying dog in a business suit that apparently resonated with viewers. A group of 200 ad execs actually preferred the AI-generated ad to a human-made version featuring a woman painting on a rooftop.

Flying dogs aside, Milo recommends that we leave the conceptualization of ad creative to the humans for the time being while using AI to predict which messaging will be successful.

“Humans are still significant, but the future will be unifying AI insights to be sure the correct ad creative is going to the right place in the right format.”

Exit mobile version