Predicting the future impact of artificial intelligence is still a fool’s errand. The public-facing technology is still in its infancy. However, marketers who don’t embrace AI solutions (even in their nascent phases) are losing out on enormous opportunities.
As we’ve previously covered, advertising platforms like Google’s Performance Max campaigns and Meta’s Advantage+ campaign rely on machine learning systems to relieve marketers of the grunt work of digital advertising. While they may still be perceived as “black boxes,” these systems have the ability to go from an intern’s understanding of where to find your audience to a master’s level in a matter of days.
And over the last several months, the public marketplace has been flooded with new AI-driven tools that can churn out work that is highly creative in nature (something once thought impossible). It has led the entire marketing industry to stand up and acknowledge how far we’ve come.
Yet, however seemingly creative these systems can be, they aren’t without their limitations.
Even so, they are powerful tools that can greatly benefit modern digital marketers. But can they benefit from the work you (yes, you) do? In order to unpack all of the potential and pitfalls of these new AI tools, we sat down with Samuel Snider-Held, senior director of technology and AI at Media.Monks, a Global Solution Partner at Funnel.
Samuel Snider-Held,
Senior director of technology and AI, Media.Monks
What does an expert envision for AI?
Sam was pretty clear from the beginning of our conversation — he doesn’t like to predict the future, but he is hyper aware of what is happening in emerging technologies. Particularly when it comes to AI, things are evolving and changing so fast that today’s prediction is tomorrow’s eye roll.
After all, it was just a year or two ago that the entire technology industry was so sure that virtual reality would be a world-changing paradigm shift, yet it never came to pass. However, for Sam, AI seems to have more traction than VR.
“This one feels real,” said Sam. “Especially as those systems that are increasingly fed by real social and search data. People are working on AI systems everywhere.”
And that seems to be the crux of what makes emerging AI technologies more real than, say, other tech-drive pies in the sky: data.
All roads lead to DCO
It is perhaps inevitable that the underlying software structures that lead to media placement optimization are now helping us improve the customer-facing creative. After all, the machine learning algorithms are already testing your ads to make sure they are reaching the best consumer at the best possible time.
Now, we are seeing new tools that can implement dynamic creative optimization, or DCO. These are technologies that can quickly create multiple versions of an ad from the same base creative, while customizing certain parts of the ad based on data fed to the algorithm.
It seems everything is coming together at a nexus that will redefine the entire industry.
“I think Advantage+ will take DCO to a generative solution,” said Sam. “At some point, we won’t even need to create assets.”
Coming from a self-avowed non-predicted, that’s quite a bold statement. However, it’s not hard to see this kind of future if you squint your eyes a bit. Tools like MidJourney can create custom artwork from the prompt of a few words. And while it may be still imperfect, it’s not hard to imagine future iterations that can create custom artwork based on your own brand standards.
Continuing to democratize the marketing space
Most modern advertising platforms now assume the role of a formerly entry-level performance marketer. They review data, determine bids, optimize for the audience, and more. This has democratized the performance marketing space. Those who didn’t have the innate ability to already optimize every available lever from the start can now rely on the algorithm to make the best decisions.
This has led organizations that are less data savvy to compete with businesses that have entire teams of experts on hand. The same may soon be true for creative executions. However, Sam doesn’t see every brand relying on off-the-shelf AI products.
“Bigger brands will still want a hands-on approach,” said Sam. “They will be looking for a sort of boutique AI creative agency that designs marketing and creative systems for their specific brand.”
The AI is not a CD
Sam was keen to point out that, no matter how advanced an AI system may seem, it is no replacement for a great creative director. It is merely evolving every creative to be closer to the CD chair.
To give an easy analogy, think about Adobe Photoshop. Back in the Ad Man days, agencies had to employ real people who were capable of editing physical photographs with actual airbrushes. Yes, real airbrushes! Nowadays, anyone with the software can airbrush at the click of a cursor with better results and less toxic fumes.
New AI technologies have the potential to do the same, but also turn artworkers around the globe into creative directors.
“With these AI tools, you’re not putting the CD into the software,” said Sam. “Rather, the CD is the person at the chair helming the ship and telling the AI what to do.
“Everyone suddenly has an army of illustrators and designers at their fingertips.”
This paradigm shift isn’t restricted to the creative team either.
“These tools can be used by the account team to better brief creatives,” said Sam.
Changing the face of pre-production
In Sam’s view, these new AI tools will streamline all sorts of tasks and make processes faster at every technical level. One such application could be the pre-production process ahead of a big budget photoshoot.
In the pre-AI world, most agencies would pitch a concept to a client with some speculative work. If approved, the agency would then need to plan the photo or video shoot to achieve the intended end result. The original speculative work could require the agency to spend money out-of-pocket using their on-hand resources, or they may even hire external talent at cost.
Those spec shots would then be dissected and analyzed to ensure the paid work would achieve similar angles, lighting and feel.
In the AI world, though, agencies can simply enter some well-thought text prompts into an imagery generator and… pow! Out comes an image that can be laid into an ad creative.
“We’re even evaluating whether this technology can outright replace some entire photoshoots,” said Sam.
Faster → better → more creative
Perhaps the most immediate impact felt by AI technologies will be the speed at which marketers and advertisers can operate. With so many processes becoming streamlined or automated, that will leave humans with more time on their hands.
That space could provide the opportunity for marketers to grow in all sorts of directions.
“It will be interesting to see what we do with this new efficiency,” said Sam. “Are we just creating more content at the same rate, though we now have more time, or are we creating new and interesting concepts now that we are unburdened of old processes?”
Humans are still required
While new advances in AI provide a myriad of possibilities, it can seem a bit daunting to marketers and creatives who once sought themselves as impervious to automation. However, according to experts like Sam, humans are still a major part of the equation.
“The ‘craft’ element of creative work must still be delivered by humans,” said Sam.
Indeed, while tools like ChatGPT and MidJourney can deliver impressive results for a machine, they can’t yet bend and break certain rules to deliver truly innovative work. And there is a reason why truly creative thinkers have been able to evade technology-driven automation and redundancies over the centuries.
Even today’s tools can’t replace the human touch. According to Sam, there is also a certain limit in how much a creative worker is willing to work with fully automated systems.
“These are highly creative people who gain fulfillment and internal validation from creating the coolest, most creative piece of art,” said Sam. “They don’t want to come to work just to push a button that does all the work for them.”
The blood, sweat, and tears are part of the journey.
What systems show the most promise?
In Sam’s perspective, the greatest promise for the current generation of AI systems are in their educational capacity. And he’s quick to point out that it’s amazing how advanced these systems have become to date.
“I’m a sci-fi nerd,” said Sam, “and seeing these systems… What they can do, literally, isn’t sci-fi anymore, and that’s so compelling.”
Sam notes the capacity of large language models to explain incredibly advanced concepts to a child.
Recounting a story of his own, Sam asked ChatGPT to explain quantum computing concepts to him and write a sample quantum computing script in the Python programming language. He then shared that code with a fellow developer.
They began to run and test the script. At first, they were a bit disappointed that the script was spitting out a few errors. However, they quickly and suddenly found themselves patching the holes and writing quantum computing code with the help of the AI, which absolutely blew Sam’s mind.
It was like going from zero to 100 in an afternoon. All powered by AI tools. Or, as Sam says it:
“Hello, year 20,000. When did we get here?”