“AI has not as of yet transformed the remuneration model,” said Forrester analyst Jay Pattissall.

Eventually it will. But for now, most of the talk is just that: talk. And when it does lead to action, it’s confined to cautious small-scale pilots — experiments that hint at the future without committing to it. There’s still too much to unpack, too many unknowns for CMOs and holding company chiefs to sign off on a full shift from billable hours to outcome-based pricing anytime soon.

As Ryan Kangisser, the chief strategy officer at media consultancy MediaSense, put it: “There is an inevitability about the shift to output models as with AI, time (which dominates so many compensation models) is often incidental to the actual output. And despite an overwhelming appetite to improve agency compensation models, there are still a few barriers to overcome.”

Those barriers tend to fall into three familiar categories: culture, technology and management. None of them are easy to dislodge. 

Realigning advertisers and agencies around flexible, AI-era workflows means moving beyond fixed headcounts and rigid team structures. Procurement teams and finance execs need clarity on what exactly they’re buying and how to measure it. Output based models demand not just delivery but definition. And in a business where “quality” can mean just about anything, that’s no small ask. 

No surprise then that these conversations rarely go beyond the pitch process — one of the few moments when CMOs have both the remit and leverage to question the model entirely.

“To work effectively, clients and agencies will need to instill clear guidance, frequent dialogue and transparent feedback to ensure every conversation doesn’t focus on fees,” said Kangisser.

What’s happening now is more of a micro-shift than a revolution, driven by select advertisers entertained by certain agencies and mostly limited to specific markets or scopes of work. Picture a localized campaign with 100 variations of the same creative, priced at $1,000 each instead of the usual patchwork of billable hours: 10% of one exec’s time, 20% of another and so on. It;s a different logic, and one that’s starting to catch on — just not at scale. Yet. 

That may change as more agencies break away from the holdco model.

Some of that change is already underway. Over the past 18 months, consultancy MediaLink has tracked a noticeable shift in how agencies think about their business models and how clients are responding. AI is already delivering measurable efficiencies in creative production, with average savings north of 27%, according to MediaLink. In some cases, asset-level costs for things like digital static images have dropped by more than 1,000%.

Performance-based compensation is also creeping into the conversation. Some agencies are floating it with select clients but tying output to outcomes remains a challenge, thanks to the industry’s favorite problem: attribution. 

As MediaLink managing director and partner at UTA Donna Sharp put it: “We’re just seeing everything is cracking from the foundation of process.”

What’s old is new again

Rumination about remuneration predates the emergence of AI in marketing. 

On the creative side of the aisle, agencies like Huge have been experimenting with output-based remuneration models for several years — the idea being that such systems granted clients transparency, and agencies fairness. 

AI is gradually turning up the heat on such efforts.

S4-owned Monks is working to reinvent the agency model before it becomes obsolete by shifting from headcount-based billing to output-driven models. This shift is laying the groundwork for a “marketing-as-a-service” model, said Wesley ter Haar, Monks co-founder and chief AI officer.

Ter Haar criticized the industry’s dependence on billable hours and constant hiring, which he says damages culture and continuity. Instead, he advocates decoupling revenue from staffing, enabling more stable, AI-powered operations. In some meetings with potential clients, he’s said Monks doesn’t expect to be their agency in a few years as much of the work being pitched will soon be automated: “We’re helping clients move from agencies to agentic.”

“Output is a huge part of the way we are briefing our teams on productivity,” ter Haar told Digiday in March. “…It takes a little while, although with a few of our biggest clients we’re getting quite close. It also opens up a different relationship between revenue and talent.”

The message is clear: in the emerging agency arms race, the advantage goes to those who productize productivity first.

Smaller agencies are also exploring

Large agencies aren’t the only ones considering new models. Kevin Kerner, founder and CEO of Austin-based Mighty & True, said AI is changing how he thinks about agency value. He expects pricing and agency fees to fall, which will allow agencies to pass along the savings faster to clients.

“We’ll have fewer people, and I think we will be much more efficient and we’ll have to pass those savings on to our customers,” Kerner said. “…We’ll still be valuable because of our ability to get to outcomes faster, but I think we’re probably going to be charging less to get those outcomes.”

As AI drives down production costs, Atlanta-based creative agency Luckie is exploring how value-based pricing can replace traditional hourly billing. When Mark Unrein became the agency’s chief AI officer, he put together a business plan for the agency’s approach to AI that estimated at least a 25% efficiency and saving and more than 70% on copywriting and other areas. 

“We’re all trying to increase that bottom line margin and get more out of each pulley,” he said.