From the first-ever banner ad to generative AI in advertising

The first-ever banner ad is a long way from generative AI in advertising. But one of the same people who worked on placing that tiny paid piece of internet history is still engaged in the adtech industry. Now, of course, he’s focused on generative AI and other emerging technologies.

The first banner ad

It was 29 years ago in 1994 when AT&T paid actual money for the very first banner ad on HotWired.com, now just Wired. Part of a crystal-ball ad campaign that foretold people working remotely from the beach on laptops, or having video conferencing meetings, the text-heavy banner ad said simply in a rainbow-colored font: “Have you ever clicked your mouse right here? You will.”

It wasn’t even called a banner ad, initially.

The first name was “tile.”

“The original concept was drawn on a whiteboard and we were drawing outlines of the websites that were out there … and we called them tiles,” says Tom Zawacki, now president of enterprise solutions at Data Axle, but formerly employed at Modem Media. “Originally, we said, if we could just take a tile and put the tile on one website and have them click there and go to another website, that would be really cool. And that was it. So that was the original original concept.”

A box, with text, linked to another website.

A fairly humble beginning for internet advertising, you might think. Pretty far from today’s ideas of generative AI in advertising.

Generative AI in advertising personalization

Now of course generative AI is the hottest tech, not the humble banner ad, and the opportunity for creativity and variety is increasing exponentially. Ad creative personalization is one big opportunity Zawacki sees in generative AI.

“One of the nice things about using generative AI is it allows us to increase the volume of production when creating copy and or visual design,” he says. “Forever we’ve been promising the delivery of personalization … what’s gotten our way is the volume and velocity of variables that create these combinations of creative message and visual design that humans just can’t create fast enough.”

I see that too, and my mind is blown by the opportunities that Amazon, for instance, has in generative AI ads. Most Amazon ads are shown on the Amazon platform, of course, which means that Amazon knows a lot about the people seeing them: purchase history, search history, maybe some of what you watch on Amazon Prime Video, maybe some of what you listen to in Prime Music, maybe what you read on Kindle.

Imagine the personalization Amazon can create — especially on-platform — in generative AI ads for the 10s of millions of products it sells. Text and image ads should be relatively easy. Video ads are also fairly doable, with a heavier compute lift.

Personalized playables are probably coming as well.

Off-platform, of course, that’s harder. With ATT, SKAN and Privacy Sandbox taking away access to device identifiers, off-platform personalization will be harder and less targeted. 

But other platforms, walled gardens, and retail media platforms ought to be able to do similar things in their own worlds: Facebook, TikTok, Snap, Pinterest, DoorDash, Uber, and others. The golden age of marketing personalization is likely going to be found in on-platform generative AI in advertising … with the possible addition, for brands, in owned spaces like apps and websites, and permissioned communications.

Generative AI to build creative campaigns

Another question: how will marketers use generative AI to build their core creative?

It’s easy, as I stated in the conversation, to want the machine to do the work. To sort of poke MidJourney with a stick and ask it to do something cool. It’s harder to make excellent, world-class, brand-compliant creative that completely fits your needs for a specific campaign or ad opportunity.

For Zawacki, the combination of biological and artificial intelligence is always going to win.

“IBM Watson did some great research in 2017, taking human intelligence, taking an artificial intelligence … and having them do a series of events,” he says. “And in every case, the combination of humans and AI working together — augmented intelligence … they called it cognitive computing at the time —  won out in all those series of tasks.”

His goal: the Tony Stark model, where you have a human intelligence directing an AI that multiplies and accelerates innovation. We’ll all probably have our own JARVIS — Just A Rather Very Intelligent System — at some point in the near future.

And that will be a game-changer.

“We are using augmented intelligence to turn our clients and our employees into superheroes,” Zawacki says.

Check out the full show

Check out the full show by watching the video above, subscribing to our YouTube channel, and subscribing to the audio podcast on your favorite podcasting platform.

Here’s a quick overview of what we all cover:

• Introduction and guest introduction
• The story of the first banner ad
• The evolution of e-commerce and social networks
• The impact of business transformation on success
• The role of AI in business transformation
• The importance of adapting to technological changes
• The future of AI in advertising
• The role of AI in mobile apps
• The power of AI in creative optimization
• The challenge of personalization in advertising
• The importance of quality data in AI
• The shift from omnichannel to omni-person