The rapid acceleration of task automation — courtesy of generative artificial intelligence tools that have emerged in recent months — has caught the attention of local advertising executives.
“I think it’s going to be a race on ways of using AI and benefiting from it,” said Kevin Joyce, partner and head of media at El Segundo-based Liquid Advertising Inc. “What keeps me up (at night) is, how quickly are my competitors going to be developing this stuff and how fast do we have to move?”
The new AI tools produce audio, code, images, text or videos that are on par with human-generated content, and have the potential to significantly reduce labor costs and raise productivity. Joyce said he sees “two frontiers” for AI at his agency. One is the creative side, where tools like Stable Diffusion, DALL-E2 and Midjourney can “accelerate the development of interesting (visual) assets that can then be redeveloped and refined.” The other is using AI to speed up the workflows for writing and editing campaign summaries.
“The timesaving is absolutely real,” he said. “Let’s say that on a weekly basis we would need to type up a report -— just summarizing the facts, not really editorializing much — might be six or seven typed pages. It’s a laborious, routine task done by entry- and mid-level staffers well. Using AI with the right guidelines so that you obscure or eliminate the confidential data … could save 10 manhours. Or taking something that somebody has written and simply asking AI to make it more concise, a process that typically would fall on a slightly more senior-level person who has a little bit better writing skills and a bit more of an eye on how to be brief, that can be done instantaneously.”
Joyce added that AI can also help with other tasks, such as generating or troubleshooting Excel formulas or developing Slackbots.
“You don’t need a coder to do that anymore,” he said. “AI is democratizing these skills that used to be prized — and rightfully so — but were bottlenecks within a workflow. You could have a bottleneck because you have one guy who is a great editor, who can really tighten up a document, another who’s that Excel person, and another who is the chart master or something like that. But now we don’t need to wait for those people, and we don’t need to overtax them.”
Media buying is an area in which AI already shines — and it has a potential to do more. Programmatic advertising uses AI to automate the buying and selling of ads on digital platforms, run by the likes of Google and Meta Platforms Inc. It can target consumers based on their location, age, gender or transactional records, but it’s usually not privy to confidential client data or other siloed information that can also inform the ad-buying.
“There’s a research tool called Global Web Index, (that has) survey data from all around the world commonly used by agencies, ourselves included,” Joyce said. “We might use that tool to establish the size of the audience that we’re targeting, the psychographics of that audience, what they’re interested in, but that tool does not exist in an AI environment right now… When (we) figure it out on how to plug that data in, in a safe and confidential way, you’re going to get in an explosive acceleration of campaign development because the user experience of putting a campaign together will go from a sequence of hands-to-keyboard research, analysis, and spreadsheets, to ‘I would like to reach gamers who play “The Witcher 3 Mass Effect” series and “Diablo” in these four countries.’ You can list as many parameters as you’d like and it will just go query the GWI data on one side, your database of historical performance benchmarks and costs on the other, and any additional factors that you may want to pull in, and give you something that may have taken 60 hours to produce before. And your job isn’t done then, but you have a massive head start, like you’re already in field-goal range at that point.”
Karl Fremont, chief executive at Brentwood-based Quigley-Simpson, is also interested in the benefits that AI can deliver, especially in analytics.
“There’s a lot that goes into taking raw data and analyzing it,” Fremont said. “We’re looking at different tools that leverage and use AI to inform at least the first set of outcomes that the data is showing us and providing information. We’re spending our time more on what are the insights that we are drawing from the data rather than actually doing the data analysis itself.”
If AI delivers on its promised capabilities, the general labor market could face significant disruption, according to the Global Economics Analyst report published in March by Goldman Sachs Group Inc.
The researchers found that “roughly two-thirds of current jobs are exposed to some degree of AI automation, and that generative AI could substitute up to one-fourth of current work,” with “particularly high exposures in administrative (46%) and legal (44%) professions and low exposures in physically-intensive professions such as construction (6%) and maintenance (4%).”
But there’s good news as well — according to the report, worker displacement from automation “has historically been offset by creation of new jobs, and the emergence of new occupations following technological innovations accounts for the vast majority of long-run employment growth. The combination of significant labor cost savings, new job creation and higher productivity for non-displaced workers raises the possibility of a productivity boom that raises economic growth substantially, although the timing of such a boom is hard to predict.”
“If I were to reach this stage in my career when the personal computer became popularized, I would need to adapt,” he said. “It’s like you either do, or you’re this weird dinosaur still using a slide rule to calculate I don’t know what. But look, did jobs go away? No. The computer was a job creator over time. I don’t think we’re all going to be sitting around. I think there will be moments when the efficiencies become really fast and we have some spare time, but this will be filled with additional clients, additional roles that will be interesting and diversify our workday in ways that will make the jobs even more satisfying.”
Jose Villa, president of Glendale-based Sensis, is also embracing the technological tide. With programmatic advertising “there was so much of AI already built into the targeting and the optimization,” he said. “Those were jobs that people used to do. You used to have to have mid- or lower-level media planners that were doing all the optimization, and now software is doing it. Those jobs don’t exist anymore. Now that’s just changed the nature of the work, but there’s still work.”
On the creative side of the agency, workload job shuffling may be deeper.
“In order for you to create an image from scratch you no longer need a studio like traditional agencies have a studio in-house,” said Pakko De La Torre-Rocha, who works as AI consultant for downtown-based CreatorUp and operates TwinChat, an AI startup in Beverly Hills. “In the old days you probably have armies of art directors, creative directors, graphic designer support, and now that’s obsolete.”
De La Torre-Rocha uses Stable Diffusion and Midjourney, which are AI-powered tools that can generate images. The move has enabled him to increase his productivity fivefold on some projects.
“A lot of people are fighting AI, and at the same time there’s a small side of the industry that is actually embracing AI because they have no choice — it’s either sink or swim,” he said. “But with AI as an ally, agencies can craft captivating narratives, ignite audacious ideas and forge novel concepts. Tailored experiences, deep audience engagement and remarkable returns on investment will follow suit. Liberated from mundane tasks, creative minds can soar to strategic heights, embracing efficiency and productivity.”