Unraveling the AI dilemma in media agencies: cost savings or revenue shake-up?

Marketers are asking: How is your agency harnessing AI and automation? Which tasks or workflows are AI-driven? How do you maintain transparency and accountability with AI use? Are there new pricing models that mirror AI-driven efficiencies? Can you share success stories where AI has notably boosted efficiency and outcomes for clients?

These queries orbit a fundamental concern: AI’s knack for streamlining and automating could significantly slash the billable hours, potentially shaking up the traditional billing model.

Well, at least to a point: After all, media agencies typically charge advertisers based on various factors such as project complexity, billable hours and the level of expertise involved. So any threat to this aspect of the model is sure to catch agencies’ attention and prompt serious consideration.

“The question of how to be compensated is a live one,” said a senior exec at a consulting firm, who exchanged anonymity for candor. “Clients are saying to agency execs ‘well if it’s going to take you 20% less time to manage our media thanks to AI, does that mean we’re going to see a 20% reduction in fees?”

Of course, there’s a lot more to this than blunt questions. Marketers are aware that AI’s capabilities in targeting, personalization, and optimizing campaigns might justify a price increase for these upgraded services, rather than a fee cut. Nevertheless, questioning costs remains prudent.

“Every discussion we have with marketing procurement teams on Gen AI,  financial savings is front and center,” said Gregg Paul, principal at consultancy R3 Worldwide. “It’s going to be the most important topic for this year. It’s up to both sides to be more transparent on how, where and when it’s being used.”

Still, don’t expect sudden upheavals. With marketers yet to take control and agencies in no position to dictate terms, the current scenario is more about ongoing dialogues, as both parties navigate the evolving landscape.

That’s why execs are cautious not to hype these initial discussions too much. They’re wary of falling into the trap of overestimating the short-term impact of new technology and underestimating its long-term effects. 

In the short term, they believe there will be little change in how advertisers pay media agencies. Looking ahead, however, agencies focused solely on selling time rather than value may face revenue losses to automation. 

In contrast, those that adapt, proving they can deliver superior outcomes per hour, may justify charging more for their strategic, high-level services.

“It may be early days still but agencies will no doubt be thinking about how to package the huge investments they’re making in AI, and how they can reallocate the time saved into more strategic, valuable services as well as maintain client revenue,” said Ryan Kangisser, the managing partner of strategy at media advisory firm Mediasense.

What’s surfaced so far is merely the beginning.

Marketers are already assessing the potential risks associated with agencies integrating AI more deeply into their operations. 

One such concern, highlighted by Mark Gay, chief client officer at Ebiquity, involves the unique advantages that agencies might gain through leveraging their proprietary data in their AI models.

“Medium term the conversation may well evolve to focus on the two-way value exchange between the data inputs a client can provide and all an agency can do with that across all their client base, and the added value the agency will return to that specific client based on how they have trained their AI — with this value equation defining the commercial model,” said Gay.

For now, however, these discussions are centered on risk, not fees, and for good reason. Marketers have witnessed the consequences of agencies accessing their proprietary data. Remember agency trading desks?

Part of the issue was the pooling of client data, providing agencies with scale but leaving advertisers feeling shortchanged. Advertisers lost control over where their ads ran, who saw them, and how their data was utilized for targeting purposes. 

It’s understandable then that advertisers would be cautious of history repeating itself with AI. 

Those concerns have pushed marketers to seek assurances on what happens to their data once it’s been fed to an agency’s AI. Is that model then being used for the benefit of other advertisers, they wonder. If not, they want to know what protections and provisions are being used to make sure this sort of risk is curbed. 

“Questions about data and IP leakage are bound to escalate as AI blurs the already fuzzy boundaries around usage rights,” said Andrew Frank, a vp analyst who sits in the Gartner Marketing Practice. “We might expect to see a resurgence of demands for category exclusivity and full-service agency consolidation.”

At its core, this is all about marketers figuring out what value their agency derives from their client relationship and whether it aligns with what they receive in return. It’s easy to see how marketers might then contemplate leveraging this insight as an opportunity to negotiate lower fees.

“We’ve got clients who are thinking about moving to a modell where maybe their agency fees reduce in return for releasing some of their data to agencies looking to build out their AI solutions,” said a exec at a global consulting firm who spoke on condition of anonymity because they were not authorized to speak to Digiday. “There are one or two CMOs we work with who are saying ‘hang on, our media agency should be paying me for this, not the other way around’.”

But agency bosses can breathe a sigh of relief, at least for now. Most advertisers are still grappling with understanding the value of their data, let alone articulating it to others.

Whatever happens, it’s going to be a slow burn, not a sudden blaze.

Yet, when these challenges do arise, those agency leaders will need to stand firm against some of these proposals — or at least not readily concede to them. That’s why it’s crucial for these agencies to proactively adjust their practices to enhance efficiency and translate those savings to clients. This has always been essential in running a successful media agency: realizing that while overall marketing budgets may not change, clients always expect agencies to find more efficient ways to deliver results.

“it comes back to the idea that as much as agencies would like to claim that they can drive performance, it’s hard, though not impossible, to prove,” said Brian Wieser, media analyst and author of the Madison and Wall newsletter. “And so in lieu of that, they have to pursue efficiency every year. And so every year, they need to have a plan in place about how they’re going to drive more costs down. You already see that in these applications of AI across the agency holding groups.”