‘It’s a death sentence’: agencies urged to urgently act on AI remuneration models

A report he co-authored ahead of Cannes outlined that ad agencies lead in AI adoption and maturity, ahead of other groups, including marketing organizations, the general business population and consumers. Its survey of American ad agencies found 75% are now using Gen AI, compared with 61% in 2024.

Yet, as it stands, the vast majority (75%) of them are absorbing the costs. “The existential threat here is that only 6% of them are able to monetize their generative AI capabilities,” Pattisall tells The Drum.

This is putting more pressure on a commercial model that’s been under mounting strain for years, if not decades. Agencies largely operate on FTE-based pricing models – charging for the time and resource spent on any one account. These models have prevailed due to their comparability in the pitch process, administrative ease and alignment with procurement. But AI is changing this, especially as headcounts decrease.

Back in 2023, Forrester forecast that 7.5% of advertising jobs in the US (around 33,000 advertising jobs) would be automated by 2030. “Based upon my observations, that forecast is conservative. It could be closer to 15-20% and we will be seeing the impacts on head count in the fourth quarter and into 2026 and 2027,” Pattisall says.

Agencies are positioning themselves as smaller, but smarter, thanks to AI. But as they change, so must their commercial model; otherwise, fewer people inside agencies will mean smaller margins and less revenue.

“It is absolutely an existential issue for agencies to be able to figure out how they get compensated for their use of AI as their head count shrinks,” Pattisall adds. “The inability to do so is, well, I’ll just say it, a death sentence to their current economic model. It will most certainly result in a number of agencies going out of business if this were to stay as it is.”

A new IPA report based on interviews with creative and media agencies, clients and intermediaries found there is “widespread acknowledgment” among all sides that these models no longer align with what’s demanded of a modern agency service. Only 27% of agencies believe they are paid a fair price for the work they do, while 58% report little to no progress in reforming commercial agreements.

“The application of AI has gotten ahead of the commercial conversation on AI,” Grey London chief executive Conrad Persons tells The Drum. “We’re at this point where there is amazing stuff that we can do creatively, strategically, efficiency-wise and effectiveness-wise. But the conversation about how we monetize that, how we alter our business model, is really fluid.”

The barriers to pricing innovation

Forrester’s Pattisall says that based on his conversations, CMOs are open to realigning the old way of remuneration to a performance-based contract or an outcome-based model. The IPA similarly says that despite marketers being hamstrung by budget uncertainty, persistent economic pressures (read: tighter budgets mean there is little appetite for taking risks), they “are absolutely ready to listen if you’ve got a demonstrably better way.”

So if agencies know they need to change and CMOs are, on paper, willing to embrace it, then why has there been so little innovation in the FTE model? There are a couple of factors, but the IPA and Forrester both landed on one common denominator: the pitch process and procurement.

The implementation of these models often has to be discussed from the onset of the relationship. Intermediaries play a critical role here. And yet during the pitch process, they often fall back on time-and-resource pricing as the only fair comparison between competing submissions. “Agencies get penalized for suggesting new pricing approaches. It’s easier to play safe and win the business,” says one agency exec.

Even if an alternative model makes it through the commercial selection stages and that agency goes on to win the business, they will inevitably find themselves in a second pitch process with a brand’s finance team to agree on an equitable contract. And it will get killed then.

“The issue around outcomes and performance-based compensation is that it’s variable and the CFO does not like variable accounting,” explains Pattisall. “They don’t want to sit on ‘unclaimed’ money that they may pay to the agency if they meet the goal or not. This is what interrupts marketers’ ability to modernize the commercial model.”

But agencies are also at fault. Jason Cobbold, chair of the IPA Commercial Leadership Group and CEO of BMB, says they get caught up in the battle to win business and fail to leave the time and space to challenge pricing strategy. “Too often, agencies let existing contractual arrangements roll over another year because change is difficult. If we are to evolve meaningfully, this has to change,” he warns.

The solution

All of this boils down to the need for a payment model that reflects the prevalence of AI in the agency proposition but offers procurement teams a clear price they are paying.

Forrester has an idea for this solution called ‘the Human Technology Equivalent,’ which essentially adds a set FTE service fee (a blended rate, which is the average of all team members’ rates on the account) with a technology fee. “I don’t think a SaaS fee is enough,” says Pattisall. “They need to couple it and integrate it with services so that the services and the software costs are linked and inseparable.”

Then there’s the deafening talk of outcome-based models, which would reward agencies based on achieving a certain set of metrics tied to a client’s business objectives. In the real world, the agencies best placed to adopt this are those that have landed an integrated account, since they control the media and creative, and pricing agreements can be struck for all the combined activity.

“I think there’s a fairly whole wholesale move away from time-and-materials and FTE models to outputs and ultimately outcomes,” says Stephan Pretorius, the man spearheading WPP’s AI platform.

“Now, the way that you get from outputs to outcomes is that you need a broader scope of influence,” he continues. “I mean, if you are only responsible for buying the paid media or producing or coming up with the original ad idea, you can’t really be responsible for the total outcome, right? So in all our commercial models now, where clients are willing to extend the scope beyond one service offering to, say, creative plus production plus media, we are able to go into outcome-based models and be very aggressive about them. So fully aligned with the client outcomes and [they] effectively underwrite our service offering.”

Grey’s Persons previously told The Drum he wants the agency to take on more end-to-end transformation briefs rather than “transactional” projects. This would pave the way for more long-term conversations on payment models.

“Are we going to switch 100% to output-based models? Probably not. Are we gonna stubbornly stick to 100% time and materials models? Definitely not. Are we going to find ways, use cases and moments of work with clients where you have an amalgamation of those things under a single roof or even within a single scope of work? Yes,” he says.

“I don’t think it’s this kind of binary decision of, is it this one or that one? But if you think about each client relationship or each agency P&L, it’s a pie. How are those slices going to get bigger and smaller over the years to come? My sense is that that pie has actually been really static, surprisingly static given all the changes happening in the world, and that will finally start to change.”

The push to bundle creative, production and media into a single service offering (with an accompanying seamless commercial model) is understandable given that the bigger advertising businesses know they will face a carve-up on their creative and production services and aggressive competition on pricing from rivals promising to do it “faster, better and cheaper” otherwise.

Among their number is BrandTech group boss David Jones, a vocal and unapologetic advocate of generative AI who has long reveled in being a thorn in the side of the holdcos.

“If you’re a traditional business that has relied on selling humans and you’ve got 100,000 of them, you’re going to have a very challenged future,” he tells us.

Jones explains that when BrandTech launched, it said it wouldn’t make TV commercials because they weren’t financially viable. “I will now say to you, we will make television commercials. Why? Because we can make them with Gen AI. So we can make a TV commercial in under a week. Let’s imagine that, all in, we’re going to make it for – let’s go crazy – $100,000. In the traditional industry, it takes six months and let’s say it costs a million dollars. Now we come along, who’ve never made TV commercials, and we have this whole new revenue stream. If you’re a traditional agency and you used to charge a million dollars, you’re now charging $100,000. Oh dear.”

Jones’s bullishness illustrates the strain that agencies will be under when their clients hear, either from him or the many other nimble upstarts eagerly disrupting the model, that there’s another, much more cost-effective way of making their marketing spend go further. If agencies don’t change how they monetize their output, change will be forced upon them.