Why Generative AI Makes Sense for Advertising

Once upon a time, in the “before times”, I was sitting at my desk, feeling a bit distracted. I was procrastinating, since I was dreading a looming project. The work I had to do wasn’t particularly engaging because the tasks were repetitive, manual, and rote. So, what did I do? Read on to find out.

But first, some context related to the headline of this article. AI has already been used in digital advertising for the better part of the 21st century, to automate bidding, upgrade targeting, and even as the first line of defense against misinformation and inappropriate content. Targeting in particular was the big “advertising automation” story of the past decade. Although we’ve seen massive signal loss from Apple iOS14 and public sector privacy changes, targeting has still been transformed by AI. What previously required manual targeting inputs (e.g. age, gender, geography, interests, behaviors, etc) is now primarily driven by automated algorithms.

The next frontier is “creative”, or the text and visual of the ad content. Creative is the last major component of digital advertising that is yet to be automated. And now, Generative AI couldn’t be a more perfect technological match for it. Even smart people without subject matter expertise recognize this, as Ezra Klein remarked on a recent podcast that “[Generative AI is]... perfect to be hooked into the online advertising schema.”

To fully appreciate why, consider first that culture is now moving faster than ever. Look at what’s trending on Tiktok week over week, or how quickly fashion evolves. If brands want to keep up, they need to refresh creatives just as quickly. But designing and developing new messaging, concepts and visual storytelling that resonates with key audiences is time consuming and expensive.

It is estimated that average production costs for TV ads range between $20k and $1M. Think about how many steps go into a TV ad production. The first stage is research to appropriately plan which messages will resonate with which target audiences. Then talent must be secured, shoots must be planned and executed. Finally, post-production is where the creative is finalized. Top notch video ads are essentially miniature film productions.

Now, imagine if you could produce essentially the same thing by only entering a short string of text as a prompt for a Generative AI text-to-video model. Sure, the quality won’t be as good at first. But it’s only a matter of time until it will compete, especially on small mobile screens. At a fraction of the production cost, it will be difficult to ignore as an option.

At this point, I imagine some of you are contemplating potential job losses in creative fields. Though I’ve been grappling with this outcome, after reflecting I’d argue the solutions will likely outweigh the problem of job losses. As writer and professor Tom Davenport argues in his book “The AI Advantage”, this revolution will hew much closer to augmentation than pure automation, and new jobs will also be created on the other side.

In this vein, Generative AI also makes sense for advertising because Generative AI augments creativity. Many people that work in marketing and advertising are “Creative” in title, function, and/or if only in desire. Tools like DALLE-2 and Stable Diffusion still require some raw creative input (text prompt), but the output allows for almost infinite net new ideation. These novel ideas can be recycled and repurposed in ways that early adopters are still just starting to figure out.

What’s more, much of the annoying, repetitive tasks are being automated away by AI, freeing up time for marketers to do higher level work that is more strategic and creative. Imagine if 50% of your job was spent doing a boring rote task like copy/pasting text or aligning elements in a design editor over and over and over. Now imagine that one day, a magic machine comes along to obviate that task, and all the sudden you have half your time back to think about new concepts or test different angles in a campaign, or heck, even just stop working earlier in the day :)

What happens in that case? You are happier in your job, so you do better work, and your organization excels because of it.

So, what did I do about those rote and repetitive tasks I teased about at the top of this post? Well, unfortunately, I put my head down and grinded them out. I wish I had an AI tool to help me automate them. More and more, we now have tools to help avert cumbersome tasks. Let’s lean into them so we have time for more strategic and creative work, and ultimately, so that we can be more effective and happier in our jobs.