Future of Marketing Briefing: The ad industry’s next mediapalooza is loading

2026 is the opening act for that moment. 

Coca-Cola’s global media, data and technology business is in play. Microsoft has already moved its media dollars. So have Adidas, IBM, Dyson, Estée Lauder, Heineken, Honda Europe, Jaguar Land Rover and Kenvue. These accounts have contributed to around $13 billion in concluded media moves this year, with a further $11 billion currently under review, per COMvergence. The full-year figure is expected to settle somewhere between $30 and $32 billion — well below the $37.5 billion recorded in 2025, the $40.5 billion 2024 and the $37 billion the year before.

But the gap is a feature not a bug because 2026 isn’t light on reviews due to dissatisfied clients so much as they’re not ready. The real volume is still loading. At least 27 of the world’s top 100 global advertisers have not put their media accounts up for competitive review in over seven years. Several others that moved in 2023 are now approaching the natural end of their contractual cycles. When those two forces collide, probably sometime in 2027, the volume of media billings in play will likely swell. 

“All clients we are working with, even those with long standing relationships, are now in active consideration,” said Ruben Schreurs, CEO of media management firm Ebiquity. 

That doesn’t mean they will all turn that consideration into full-blown reviews. Top-to-top conversations can achieve meaningful change if both sides are honest about how the market has shifted since the contract was signed. But reviews remain the sharper tool because the entire industry is reshaping itself at once, and the briefs that come out of that reshaping will cut across media, production and creative in ways that existing relationships, however good, weren’t built to absorb.

Which makes 2026 Infinity War to 2027’s Endgame. The accounts moving now are the scene-setting — the positions being taken, the holdco relationships being tested and the agencies figuring out which version of themselves they need to be before the real reckoning arrives. Whether that reckoning changes anything is another question. It never quite has before. 

Then again, those previous mediapaloozas had a single animating question. 

In 2015 it was whether agencies were spending honestly. In 2018 it was whether they had the tech capability. In 2020 it was whether they should exist at all. This one doesn’t have a single question. It has two sides that, for the first time, want similar things. Clients need contracts that reflect a different industry, one where AI governance and principal media are table stakes rather than afterthoughts. Agencies need a commercial model that isn’t slowly killing them. That overlap doesn’t make the negotiation easy — they still disagree on the terms — but it does make this round less antagonistic than any that came before it. 

The evidence is already showing up in the contracts themselves. Newer agreements between advertisers and agencies include disclosure clauses on how much money is being made through principal media, spelling out when an agency can act as principal, in which channels and under what conditions. 

The same goes for AI, though the contractual language around it is still being invented in real time. The immediate pressure point is cost. Agencies have spent the last two years absorbing AI infrastructure and carrying it on their balance sheets as a client acquisition tool. That subsidisation is ending. AI costs are starting to move off agency balance sheets and onto client ones, and the contract negotiations now reflect it. Some agencies are offering to continue absorbing AI infrastructure entirely, but only if the client commits to a minimum principal media allocation in return. Others are charging for token usage directly, with a premium, in the same way overheads were historically applied to headcount. The structure differs deal to deal. What’s consistent is the direction. 

“Just saying ‘we’re doing better work with AI’ is not good enough. You need to explain what, why,” said Philippe Dominois, co-founder and CEO of Abintus Consulting. “It will be quicker and cheaper, but actually is it effective? You need to find the right balance between the two.”

Those are questions a pitch forces agencies to answer in a way a mid-contract conversation rarely does. Which is why, for all the talk of transformation, the accounts going into review in 2027 may offer the industry its clearest shot yet at resolving the one problem it has been deferring since 2015: how to pay agencies for what they actually produce rather than the hours they spend producing it.

Every mediapalooza has promised that shift. None has delivered it. The obstacle has never been resistance so much as infrastructure. Tying media performance to business outcomes is harder than it sounds when pricing decisions, competitive moves and macro conditions can move a brand’s numbers regardless of how good the work was. What’s different now is that AI has made the inputs so cheap and so fast that paying for them by the hour has become indefensible even to the agencies. The old excuse — that measuring output was too difficult, so billing for input was the only workable model — is running out of road. The next wave of reviews will test whether the industry can finally build something better, or whether it reaches for the same answer it always has and calls it new.

“If 2021 coming out of covid was very much about efficiency and simplification, I’d say 2024 was recalibration — and I think 2027 would be far more transformative,” said Ryan Kangisser, chief strategy officer at MediaSense, the consultancy managing Coca-Cola’s current review. “They want transformation, they don’t just want another media agency.

What that looks like in practice is still being worked out. But the direction is clear enough: something interoperable, where a core agency relationship handles the bulk of the work but the client retains the right to bring in specialists from outside the group without friction or penalty. The holdco answer to that has always been plenty of choice, all of it ours. Clients are starting to find that unconvincing. But the independents and specialists they’d turn to instead aren’t yet big enough to absorb the volume. So for now most will design something that covers the majority of their needs through the main relationship and builds flexibility around the edges. The fight is to get that flexibility written into the contract. Most current ones don’t provide for it.

“Clients are still treating pitching as the default answer to problems it rarely solves,” said Patrick Ryan, Founder of The 300 Consultancy. “If moving agencies was the fix, we wouldn’t keep seeing the same accounts come up for review so regularly, and awarding a different agency partner each time.”

Short-term thinking, he continued,  only amplifies the problem tied with high turnover on both sides means relationships are constantly being rebuilt instead of strengthened.

“If we want a more progressive industry, we should all be striving for better partnerships, ones that are built to deliver results and withstand change together,” said Ryan.

Outcome-based remuneration is loading, too

No quest in advertising is more quixotic than outcome-based remuneration. It has looked close for years. A deal here, a soundbite there, then a mirage. Because tying media performance to financial performance is genuinely hard. There’s too much outside media that shapes a brand’s results. Still, the industry keeps chasing it, anyway. The reward is too big to give up.

But something is different about this latest push. The holdcos are bullish on it, albeit self-interested given their current plight. CMOs are eager too, even with procurement standing in the way. Then there’s AI, forcing both sides to reckon with what a new commercial arrangement actually looks like. There’s a confluence of forces, in other words, making the discourse around outcome-based remuneration more tangible than it has ever been.

One of this year’s most prominent media deals is rooted in exactly that. Jaguar Land Rover has agreed to tie a substantial share of its fees to WPP to measurable sales and brand performance, rather than hours worked. Conversations are already happening, too, over how to reprice the unit of work itself, an adjacent part of the same broader remuneration debate. Ripples, not tidal waves. But movement all the same.

“Outcome-based remuneration is real,” said Ruben Schreurs, CEO of media management firm Ebiquity. “The type of deal making and the type of skin in the game from agencies and clients is changing, because there is this context momentum around it, accelerated by AI and adoption of AI in all facets of the marketing operations.”

Granted, it will play out, like these things always do, along a spectrum. Some, like Jaguar Land Rover, are already there. Others will get there eventually, albeit after a lot of change. Because making outcome-based remuneration work isn’t as simple as agreeing to pay an agency for sales. There are too many variables outside an agency’s control for that to be fair to anyone. A pandemic hits, a competitor cuts price, a hot summer does an agency’s job for it. None of that is the agency’s doing, and none of it should be scored as if it were.

But even when the model works as intended, there’s a harder ask underneath it: CMOs have to be secure enough to watch their agency’s fees rise and read that as a good sign, not a bad one. An agency going from $5 million to $10 million in fees should be a no-brainer for many CMOs so long as it’s tied to the incremental profit the client wouldn’t have had otherwise. That way, the fee only looks big in isolation. Against the number it’s supposed to be paid out of, it’s cheap.

This is where media mix modeling (MMM) comes in, or rather the latest iteration of it.

For most of MMM’s history, that kind of trust was expensive to earn. A single study took six months and at least $250,000, which meant only the largest advertisers could afford to run one, and even then, once a year at best. Compute and automation have cut that cost roughly in half and the turnaround to as little as four to six weeks. Clients who used to treat annual modeling as a stretch are now asking for it quarterly.

That speed matters because it’s the thing that lets MMM do the job neither side can do for itself: separate what the agency actually controls from what it doesn’t. Mix models can mathematically strip out the noise a raw sales number can’t, pricing changes, a competitor’s move, macroeconomic drag, so that what’s left is closer to media’s true contribution. A lift model won’t catch a competitor raising prices 10% during a campaign. A mix model will, and it will tell you the media did fine, the market did the rest. Run quarterly instead of annually, and at the campaign level rather than the channel level, it can also show a brand that one campaign returned 10 times what another did — the kind of granularity that used to take a year to surface and now shows up in weeks.

And yet, none of this guarantees the path to outcome-based remuneration is any easier this time. There’s still a fear it reverts to the mean, becoming something the industry keeps talking about but never arrives at. Not least because the debate itself has become a business, and that works against anyone actually landing on an answer.

“It’s a bit like Waiting for Godot in that it never actually comes, the industry never actually gets to that point,” said Ian Whittaker, managing director of Liberty Sky Advisors.

On that basis, marketers would be wise not to wait for outcome-based remuneration to become a scientific, auditable formula. It may never come, and some people are financially motivated to keep it that way. The realistic version, according to Whittaker, looks more like a negotiated value judgment than a mechanical payout. The industry’s actual failure, in his view, is that it’s stopped trusting itself to make that argument confidently.

What we’ve heard

“Typically media as a creative production is based on an hourly rate. How long can you keep that FTE when you’re using tokens? How do you reprice creative in a token-based economy? Should, for instance, agencies become futures markets for tokens, where if you assume they are becoming more expensive in the future, they should be holding them and licensing them Perplexity-style to customers.”

—Daniel Knapp, chief economist at IAB Europe

Numbers to know

42%: Percentage of enterprises that abandoned most of their AI initiatives in 2025

€34 billion: Total video advertising revenue generated in Europe in 2025, and accounted for more than half of all display investment for the first time

38.2%: Percentage of U.S. online shoppers that haven’t used a retail AI chatbot and aren’t interested in doing so

42.4%: Percentage of U.S. adults that believe brands should remain neutral and focus on their products/services rather than get involved in politics

What we’ve covered

OpenAI set to expand ads to France, Germany and Ireland
OpenAI is hiring ad execs in Paris, Munich and Dublin, signalling an expansion of its advertising pilot into France, Germany and Ireland — less than three weeks after its U.K. launch. With parallel hires in Singapore, the pilot would reach 10 markets, though OpenAI is still only asking advertisers for test budgets.

“Brands recognize the cultural dominance of podcasts”: global podcaster Mel Robbins talks AI, ad budgets and audience ownership
At her first-ever Cannes Lions, Mel Robbins argued podcasting has been wrongly bucketed with radio in media budgets and is only now being recognised for its cultural and commercial dominance. She also told Digiday why owning your audience beats any platform, and why she won’t write another book.

What’s really driving Europe’s €131 billion ad boom
European ad spend rose 10.5% to €131 billion, with digital taking around 70% of the total. Video is the engine — now more than half of all display investment — while social climbed 19.2% to €35.5 billion. Retail media grew 16.7% to €13.3 billion, though that’s mostly mediafied trade budgets rather than new money.

Advertising’s confidence problem, according to the man who used to price its stocks
Former City equity analyst Ian Whittaker told an IAB Europe panel that advertising has “lost its way,” talking to itself rather than the businesses paying for it. His diagnosis: holdcos built around media buying because it scaled, throwing in creative for free and giving away the one thing that couldn’t be commoditized on price.

What we’re reading

Denmark steps into an EU court fight over what platforms owe publishers
Denmark has intervened at the CJEU to back Belgium against Google, Meta, Spotify and Sony, who say Belgium’s enforcement of the press publishers’ right exceeds EU law, according to The Next Web. The ruling will set the limits on how far any country can force platforms to pay.

TikTok launches Discover America tour
TikTok’s “Discover America” is a six-city US roadshow tied to the 250th anniversary, spotlighting local creators and businesses while pitching marketers with reach data and trending-search stats. A post-ownership-change charm offensive, per Social Media Today.

Musk’s SpaceXAI releases new model, Grok 4.5
SpaceXAI’s Grok 4.5 targets coding and agentic work at $2/$6 per million tokens, undercutting Anthropic’s Opus 4.8. Musk claims “Opus-class”; Axios said it beats mid-tier, not frontier. Trained on compute, SpaceXAI also leases to rivals.

How social media is changing Wimbledon from eminent tournament to ‘bucket list’ event
Wimbledon’s social media hype has made it a tourist attraction as much as a tournament — influencers trading on the strawberries-and-cream aesthetic, brands flying in creators who’ve never watched live tennis. The club courted it. Courtside, fans capture the moment rather than being in it, per The Guardian.