A recent survey aimed at shedding light on AI and its applications in marketing and agency work revealed that the more people know about AI, the less convinced they are that it will replace humans.
That's one of the findings of a joint study by Omnicom Media Group agency PHD Worldwide and WARC, which surveyed 700 marketing and agency executives across 10 countries, including the US, UK, Australia, China and Brazil, between May and June this year to better understand how they plan to incorporate AI into their operations between now and the end of the 2020s.
The research is part of the media agency’s wider effort to become a hub for learning and experimentation around generative AI that is accessible to all Omnicom clients (not just PHD and OMG) and the public.
To that end, PHD is launching Ascension this week, what it is calling an AI-powered “generative publishing platform,” according to the media agency’s global CEO Guy Marks and global head of strategy Mark Holden, who said the cost will be “easily six figures” but declined to share details.
Marks said he wants to avoid the launch of Ascension being seen as just another headline-grabbing hype or nonsense, and explained what he wants customers and the public to take away from the effort.
“Essentially, there's two speeds,” Marks says. “How do we frame that narrative and communicate to our clients, our employees and the broader industry our view of where we're heading and how everyone needs to prepare for that? And then there's the question of what we're going to do in the next 30, 60, 90 days and how do we prepare not just our own talent but our client organizations to take advantage of the benefits that are in front of us?”
Holden likened the emergence of generative AI to “a kind of Cambrian explosion moment – where everything changes.” “Our view is that in a market where there's rapid change that's impacting your customers and of course your employees, in some sense it's your duty if you're an agency or any form of supplier to put information out there to educate people and keep them updated. If you don't, people won't be able to keep track of developments and as a result you lose connection with your customers.”
The research with WARC helped shape the editorial mission and perspective of the “magazine,” which analyzes the remainder of the decade into three eras of AI evolution.
The AI Experimentation Era (2024-2026) will be characterized by the separate use of large language and diffusion models (used by Ascension) and the early integration of generative AI into enterprise platforms. It will be a period of pilot projects and refinements, before moving to the next stage. The AI Acceleration Era (2026-2028) will see AI “become a fundamental element of marketing, enabling more mature applications of generative AI in existing platforms, primarily for efficiency and effectiveness, but used by a large portion of the workforce,” Ascension said. This era, with its focus on replacing or enhancing current capabilities, will see…
According to the study, the AI-advanced era (2028-2030) will focus on more intelligent systems that can make more reliable decisions, allowing workers to dream up “new marketing capabilities beyond what can be imagined today.”
Greg Stuart, CEO of MMA Global, a trade group that does deep dives into AI experimentation and testing for marketers and ad tech players, said he believes agencies' work on AI and new technology developments often seems deep on the surface but doesn't delve much below the surface.
“What I want to ask is, what frameworks are you advising your clients to look for opportunities within AI,” Stuart asked. “What opportunities are there in those frameworks that marketers need to dig into? And what knowledge do they have now that no one else has? If[agency]can't answer those questions, they can come off as superficial.”
What did the study find?
One thing to note is the gap between the desire to adopt generative AI and how well it's actually being used: While 35% of marketers feel they should be using generative AI at a high or very high level, only 27% are using it at that level, compared to 36% and 26% of agency respondents, respectively.
The survey also included a quiz to measure the gap between what respondents say they know and what they actually know. While 42% of marketers consider their knowledge of generative AI to be advanced, only 13.7% of respondents correctly answered at least two of the five questions in the quiz.
Finally, as the knowledge base about generative AI grows, more respondents understand that human participation will continue to be necessary for it to be successful: only 21% of highly knowledgeable respondents (i.e., those who answered at least two of the five quiz questions) believe that generative AI will replace practical tasks currently performed by their institution in the near term, compared to 46% of respondents with average or below average knowledge.
“The more I learn about generational AI, the more I think, 'Wait a minute,'” Holden says. “There's no big green button. It's not about replacing functions, it's just about replacing roles with other functions.”
MMA's Stewart said the PHD/WARC findings highlight a gap that will soon separate marketers from companies that get AI faster than others. “AI is going to overwhelm us all, so we need to get this figured out fast,” Stewart said. “It's like seeing a tornado in the distance and deciding in that moment that you want to start working on safety, when you should have done that two or three years ago.”
