By Michael Bürgi  •  August 11, 2025  •

Ivy Liu

So with all this newfound skill, executional ability and gained efficiencies — which have enabled media agencies to reduce personnel and save thousands of working hours — at what point are clients looking for payback? If an agency can execute at a higher level and save money, shouldn’t the client benefit financially? Certainly any procurement exec or brand CFO will believe so. 

More than anything, the impact of generative AI on remuneration between agency and client is leading both down the road toward outcomes — a road many in the agency world believe should have already been traveled. 

“Frankly, if you believe you have the luxury of waiting for clients to ask about AI-driven compensation models, you might already be behind the curve,” said Florian Adamski, global CEO of Omnicom Media Group. “There are two types of clients — the ones that will ask and the ones that will take matters into their own hands. As an agency leader you’re not in a great position if you don’t start the conversation with either category as there might be a tendency to start viewing agency outputs as commodities, not strategic value.” 

A move toward outcomes also makes it harder for procurement to find places to shave back remit to agencies. “We’re moving to more value pricing,” said Mark Penn, chairman and CEO of Stagwell. “It’s not as easy for procurement to break down the various steps into how many people hours is everything? Because [the costs involved in work for clients] is a unique blend … now there are many different things that go into the baking of your media cake.”

This is the direction most agencies investing in generative AI should be heading in — but by no means is this an easy road to travel, as decades of habits aren’t easy to break. “Not a lot of progress made from a media perspective on [moving toward outcomes compensation],” said Cyd Falkson, svp of strategic accounts at consultancy MediaSense. “Agencies deserve to be paid for the value of the work … There’s a lot of intention around moving to an outcomes based model, and yet it’s really hard because there is so much other noise that gets tied up in what drives an outcome.”

To Adamski’s thinking, that problem can be solved — in fact it needs to be, given the pressures on clients to show growth. “Clients feel the cost pressure every single day. AI is more than just a cost cutter – it can help to reallocate investment to where it drives incremental results which is a win for both sides,” he said. “Preparing clients for value-based conversations rather than time-spent ones is what creates an opportunity to evolve into a higher-value partner … Being open to explore outcomes-based compensation models now is a sound approach.”

Some clients may choose to revisit fees paid to their agencies with more of a surgical bent. Since gen AI has already shown it save time and effort in content production and research costs, as well as enabling a massive increase in creative output, those areas offer the best chance for trimming back costs.

“Marketers are not necessarily looking to remove the costs out of the fee, per se — they’re very much sticking to the line of AI as an augmentation technology,” said Jay Pattisall, vp and senior agency analyst at Forrester. “They’re not really looking to take the margin out of their agency’s fees. What they’re looking to do is remove some of the hard costs that are associated with their content and production, some media elements of it … Some companies will choose to reinvest the savings that they have into additional media, or they will choose to reinvest the some of the production savings into velocity of content.”

One innovation executive at another holding company who’s been building new AI-driven research modules, argued that clients who do push for savings over redeployment of assets are missing AI’s potential in elevating quality of work.

“If you have a finite amount of time, don’t you want the majority of it to be deep thinking?”asked the exec, who declined to speak for attribution in order to avoid angering clients. “Wouldn’t you rather us spend time on this vs. cutting time?  I do think that some of the industry needs to hear is that it’s not a magic bullet. Number one, you need more people to understand the infrastructure of how to use it, and Number two, analyze it the same and have deeper thinking and have critical thought around it.” 

It’s inevitable that agency headcount will ultimately be reduced as a result of AI’s ability to quicken execution and think on its own (within reason). “We might see a shift in how we scale and up-level our teams,” said the innovation exec, but he added that cutting back doesn’t mean instant savings for the client. “It’s not about client A reads this report from me, and then all of a sudden they’re like, ‘Hey, we got to cut scope by 10% because you’re using AI.’ I would say that’s actually the wrong thinking from a marketer to understand the impact of this.”

As MediaSense’s Falkson summed up: “In order for the ecosystem to remain alive and healthy, we have to have healthy agencies, and pushing them to squeeze ever tighter doesn’t help sustain the ecosystem.”

Color by numbers

For better or worse, sustainability in media has lost a bit of its luster, pushed aside by a governmental agenda that seems to actively want to roll back progress made on environmental issues affecting the planet. Still, there are pockets of activity, one of them is a case study conducted between ad-tech platform Equativ and independent agency Butler/Till, which executed a digital media effort for an unnamed insurance client with the aim to reduce carbon emissions — without sacrificing performance metrics. The results:

• 86% video completion rate (+239% vs. standard PMP)
• 0.20% Click-Through Rate (+18%)
• 90% Viewability (+15%)
• 17% reduction in CO₂ emissions
• Carbon emissions saved with the campaign were equivalent to nearly 20,000 miles of driving, or six round trip flights from Boston to London.  

Takeoff & landing

• Omnicom‘s acquisition of Interpublic Group cleared yet another major hurdle last week when the U.K.’s Competition and Markets Authority blessed the impending union. The deal is expected to close this fall, barring unforeseen impediments by the U.S. government. 
• WPP Media and Nielsen struck a new agreement that gives the holding company audience measurement tools covering TV, streaming, audio, and cross-platform ad performance, all through the rating’s giant’s Nielsen ONE platform. The data will flow into WPP’s Open Media Studio. 
• Account moves: Omnicom Media Group’s PHD landed global media planning and buying duties for OpenAI, expanding on campaign work it had done for the creator of ChatGPT … Tinuiti landed media AOR duties for Ancestry, which represents a bit of a return for the brand to the shop since Tinuiti handed performance media for it in 2023. 
• Personnel moves: Omnicom Media Group’s OMD promoted Ellen Griffin to be its global COO, up from chief client solutions officer … Digital media shop Coegi, owned by True Independent Holdings, named Tessa Ohlendorf its president for North America … Exverus Media hired Josh Edelman to be its vp of analytics, coming over from the same position at Giant Spoon … B2B marketing agency Transmission hired Josh Neland as its vp of AI. 

Direct quote

“Upfront season may be focused on Nielsen’s Big Data + Panel solution, but let’s not mistake counting for impact … Some alternative providers have de-emphasized currency entirely, and the buzz is fading. What matters isn’t who counts the most — it’s who proves outcomes. Marketers need to understand how media drives brand, behavior and business results.”

— Chris Kelly, CEO of alternative measurement firm Upwave.

Speed reading

• Sam Bradley looked at the impact of generative AI on the creative side of the agency business, and how to push back on the expected cost savings from clients. 
• On a similar theme, Bradley and Seb Joseph examined AI’s ability to operationalize the creative side of the holding company world. 
• Kimeko McCoy explains how the popularity of influencer marketing has unintentionally led it to fall somewhere between search and social —but how that’s also led to new opportunities for brand discovery.