As advertising AI race heats up, agencies scramble to find their USP

From on-stage lectures to rosé-fueled demos in cabanas, all the major agencies jostled to prove they were leading the way in deciphering the technology. After months of negotiating access agreements with the likes of Adobe, Nvidia and OpenAI and building their own client-facing consoles, Cannes 2024 was go-to-market time. The problem, from this vantage point at least, was that many of these ‘bespoke’ agency offerings looked uncannily alike, relying as they do on the same core AI models provided by third parties.

Though we are still in the early days of adoption, evidenced by there being no paradigm-shifting AI work stealing the show in the Palais, competition among ad agencies to prove their readiness is fierce. According to Forrester’s new report on the state of generative AI in US ad agencies, 61% of respondents reported that their agency is already using genAI, with a further 30% saying that their agency is currently exploring genAI use cases. Per Forrester, “this adoption rate stands well above other groups, including marketing organizations, the general business population and consumers.”

It is the bigger players in adland that are most eagerly embracing the technology. All respondents at large agencies (201+ employees) report that their agency is at least exploring genAI use cases, with 78% already using the technology. And creative agencies are leading the way, with 69% of respondents from that field reporting they currently use genAI versus 57% within media agencies.

All of this adds up to an increasingly crowded space for creative shops. As Accenture Song managing director Nevine El-Warraky said during a press show-and-tell on the firm’s yacht in Cannes: “The biggest question I get asked quite regularly is, in a world where everybody has access to the tools, how do we build competitive differentiation?” El-Warraky was referring to clients, but the same question could just as easily be leveled at agencies like Song.

AI models transform agency models

Out of fear of being left behind and in the expectancy of clients’ curiosity, agencies have built out their AI practices at remarkable speed. After all, what advertiser could possibly resist the technology’s fabled promise to exponentially scale creativity while simultaneously slashing the costs involved in doing so?

If only it were quite that simple, as those at the coalface would tell you. First, there is the significant upfront cost of adding capability. Then, there’s the internal upheaval of actually putting it to good use.

“My main plea to help the wider industry benefit is to stop talking about short-term efficiency gains, cost-cutting and the narrative that AI is free,” says Alex Dalman, who set up VCCP’s dedicated AI agency Faith a year ago. “There will be efficiency in the long term. However, I’m afraid to say there will need to be budget investment upfront for a lot of the model building and training, not to mention for brand data sorting and cleansing.”

In Cannes, Dalman was joined by Simon Valcarcel, marketing director of client Virgin Media O2, to exhibit some of their early experiments with the tech. These included the O2 Bubl Generator and O2 Copy Checker, designed to remove some of the barriers in complex 3D asset production and controlling tone of voice. The impression was of agency and brand on a mutual journey of discovery.

“It’s just not a ‘click of your fingers and it is done for your brand’ type solution yet,” says Dalman. “This is not a popular opinion at the moment, but it is the reality: somebody has to pay for this – the technology, the skills needed to use the technology, the cost of generation, and ultimately for the wider advertising industry to reap the benefits from it, brands will have to pay for it.”

For the last year, advertising’s biggest holding companies have been making big bets that they will. WPP announced a £250m-a-year investment in AI; Publicis Groupe €300m over three years; Accenture, parent to Accenture Song, $3bn over three years, albeit across its entire consultancy business. Privately, or at least within earshot of industry journalists, it has become an increasingly popular pastime of agency leaders to express dubiousness at their rivals’ declared spending.

Whatever the precise amounts, as Forrester principal analyst Jay Pattisall says: “The costs to build these models are tremendous.” That’s why he doesn’t subscribe to the notion that clients might seize the AI revolution as an opportunity to cut out the middleman altogether and move the bulk of their creative production in-house. “Agencies, with their buying power, are able to get access to the computing power of major players like Adobe, Nvidia, Google and Microsoft. Add to that the cost of engineering, development and training and most companies won’t be interested in that type of investment when it can be bought from an agency.”

But will agencies’ competitor set remain other agencies? Proprietors of AI tools, such as Adobe with its popular Firefly suite, have positioned themselves as agency allies while keeping the door open to direct brand relationships. In conversation with The Drum at the Adobe Summit in Las Vegas in April, Don Bennion, senior director of product marketing for Adobe GenStudio, made no secret of the fact that it is approaching marketers at America’s largest brand names. “We’re targeting enterprise… as big as it gets,” he said. “Coca-Cola? Absolutely.”

In Cannes, the mood between agencies and their AI brethren was one of fraternity once more, with Adobe and S4 Capital agency Media.Monks announcing a fresh tie-up during an episode of The Drum Show. They claim the combination of GenStudio and the Monks.Flow platform will “help marketers reduce costs [and] advance creativity.”

Nvidia, the artificial intelligence computing behemoth that is now a year into a major partnership with WPP, was also on the charm offensive in Cannes, with its global agencies and advertising lead Jamie Allan praising the creativity of the ad industry while setting out the firm’s ‘agnostic’ position on partners. “We are the only AI company in the world that works with every other AI company,” he says for emphasis.

Diplomatically swerving the suggestion that agencies might feel vulnerable about AI enterprises eating into their client base, Allan defines Nvidia’s role in the ecosystem as being that of an “enabler” rather than a crusader. “What you see at an event like Cannes isn’t particularly going to change,” he says. “We’re still going to have organizations speaking to their suppliers, speaking to their clients about how they can leverage technology to make the most impact and create the most value. That fundamental hasn’t changed, the technology will change, the speed at which we can create new capability will change, because we’re moving into this era of AI factories creating intelligence, which creates value much quicker than before.”

Humans – not tech – set agencies apart

Based on the assumption that most major agencies will have a similar suite of tools at their disposal, AI firepower alone will not be a sufficient agency differentiator. What will make them stand apart is how imaginatively they can use their new arsenal.

“Every agency has a unique selling point with genAI that comes from the unique creativity of their talent,” says Forrester’s Pattisall. “AI doesn’t create advertising. Talented creatives using AI create advertising. And that unique talent will become important distinction for agencies in the coming months.”

While much has been written about AI threatening creative talent, it is the same supposedly endangered species that agencies are entrusting right now to give their own brands the edge. Take Accenture Song, which put chief exec David Droga, perhaps advertising’s most famous working creative, front and center at its AI pitch to journalists in Cannes. Speaking with his arm in a sling from a tennis accident (write your own jokes here about human limitations), Droga stresses his belief that individual creatives can make all the difference at this inflection point for adland.

Droga, pictured above in full flight presenting before breaking off for a quieter conversation, tells The Drum: “I think in the beginning, the speed and the excitement and the momentum of force will squeeze a lot of creativity at the margins. But I think it’ll make a comeback.

“When every client has access so they can make their own content, they are really excited thinking, ‘I never have to pay a translation agency, I can have this thing write my social posts, I don’t have to pay a photographer any more.’ And to a large degree, there’s a lot of that stuff that can be done. But there are two classes when I think of messaging or consumer advertising – there’s great and there’s straight. AI, with time, will allow us to make things and operate. But there are times when you need to make connections, you need to be distinctive, you need to stand out. Taste and ridiculous things like that, which are still imperative and exist in the world, are gonna be important. I have to believe that.”

For Droga, who also sits on parent firm Accenture’s global executive management committee, this is the time to champion creativity as much as technology. “I am just advocating that there are many creative voices upstream participating, as opposed to waiting downstream to see some of the ripple effects,” he says. “A creative person who gives a shit has to be in a position of power and influence.”