Loading...
Coca-Cola's $1.7 billion WPP versus Publicis review is an infrastructure decision. The strategic accountability layer it should be testing remains unclaimed.
The largest agency showdown of 2026 is live. On June 3, Coca-Cola opened a global media, data, and technology review pitting its two incumbents — WPP and Publicis — against each other for roughly $1.7 billion in non-North-America spend, with the full account valued at more than $4 billion once Japan, Korea, and the rest of the global footprint are included. Coca-Cola confirmed the review to Ad Age and The Drum, framing its stated rationale as evolving a "digital-first marketing operating system for future growth" with explicit reference to "agentic AI tools." WPP Open X has been the global network partner since 2021; Publicis already won the North American media mandate in March 2025. The review concludes in the fall. The whole agency trade press is treating it as a referendum on the WPP holdco model.
The question the trade press is asking is the wrong one.
The question being asked is: which platform — WPP Open X or Publicis Groupe — is better equipped to run Coca-Cola's marketing operating system? The question that determines whether the next decade of marketing actually produces strategic outcomes is: who is accountable for the strategic decisions that the operating system is built to serve? One is an infrastructure question. The other is an accountability question. The $1.7 billion is being spent on the first. The strategic value is in the second.
The buyer is making an execution-layer decision while the strategy layer remains unclaimed.
The review matters less for the specific outcome and more for what it reveals about how the largest buyers in the world are framing the AI transition. MediaPost's coverage makes the operating-system framing explicit: the review is about media, data, and technology infrastructure. WPP retains creative and PR, which are the parts of the work that touch the strategic question. The review is asking which bidder can run the pipes, the measurement stack, and the data flows. It is not asking which bidder will be accountable for the strategic decisions Coca-Cola's leadership has to make about which markets to compete in, which consumer segments to prioritize, and how the brand should be positioned as the carbonated-soft-drink category is being reshaped by GLP-1 adoption, regulatory pressure, and functional-beverage entrants.
That second question — the strategic accountability question — is the one that determines whether the $1.7 billion in operating-system investment actually moves Coca-Cola's P&L. And nobody in the review is being asked to answer it.
This pattern is not unique to Coca-Cola. The same execution-layer framing is showing up across the market. According to Gartner's 2025 forecast, more than 40% of agentic AI projects will be canceled by the end of 2027 — not because the technology doesn't work, but because the organizational readiness to translate agentic AI into accountable decisions isn't there. Boston Consulting Group's Build for the Future 2025 report found that only 5% of more than 1,250 global firms are generating real value from AI at scale, and the gap between those 5% and the rest is widening. The bottleneck is not infrastructure. The bottleneck is decision architecture — and that is the layer the Coca-Cola review, like most large-enterprise AI pitches, never touches.
The dominant industry frame right now is "human + AI" — the assertion that the winning model pairs human judgment with AI execution. It is the frame the Cannes Lions festival will reinforce for ten days in June. It is the frame WPP, VML, the new Cannes AI Craft subcategory, the publicis/LiveRamp infrastructure pitch, and the major platform vendors all use. City AM reported this week that Martin Sorrell publicly called WPP "catatonic" at SXSW London the same week Goldman Sachs issued a Sell note on the stock with a 240p price target — projecting free cash flow of just £684m by 2028, versus £1 billion-plus in 2022 and 2023, and warning that WPP's planned £500m cost-savings would likely be absorbed by staff costs. The story the financial markets are telling about the "human + AI" defense is that it is increasingly defensive.
"Human + AI" is a necessary condition for any AI-augmented work. It is not a strategy. The question is not whether humans and AI work together — of course they do. The question is whether the firm in front of you is accountable for the strategic decisions your business needs to make, or just for the work informed by their judgment about those decisions. Every major AI services pitch in 2026 will tell you they combine human judgment with AI. Almost none of them will accept accountability for the decisions that judgment produces. That gap is the strategic accountability gap. It is the layer the operating-system reviews do not reach.
When a buyer hears "AI strategy agency" in 2026, they are hearing one of three very different things, and the differences matter.
The first is the infrastructure layer. These are the AI platforms, data infrastructure vendors, and operating-system providers — Publicis's LiveRamp data bet, the WPP Open Pro direct-to-brand platform, the agentic AI tooling from Google and Microsoft, the CI tools that produce dashboards and signals. They sell access. They require teams to operate. They produce fragments of insight that never connect unless someone in your organization takes the time to connect them.
The second is the execution layer. These are the AI agencies and creative shops that run campaigns, produce assets, optimize media, and deliver work. They are excellent at the work they do. They are not accountable for which work your business should be doing in the first place.
The third is the decision layer. This is where the strategic accountability lives — where the question of "given everything we know about our category, our competitors, our consumers, and our capabilities, what should we actually do" gets answered by a team that owns the recommendation and is accountable for the outcome. The decision layer is the layer that turns the infrastructure and the execution into P&L movement.
Most AI pitches in 2026 are competing in the first two layers. The third layer is largely unclaimed. The Coca-Cola review is asking the first-layer question with a $1.7 billion budget. The strategic question — the one that will determine whether that $1.7 billion produces a return — lives in the third layer, where the buyer has not yet made anyone accountable.
Cannes Lions opens on June 22. For ten days, every vendor, agency, and platform in the AI marketing category will pitch the same "human + AI" frame. The pitches will sound different on the surface. Underneath, they will all be selling the same first-layer answer to a third-layer question.
The test for any pitch — whether the bidder is WPP, Publicis, a CI tool vendor, a creative AOR, or a self-described AI agency — is whether the firm accepting the engagement is also accepting accountability for the strategic decisions the engagement is supposed to inform. If the answer is no, the buyer is buying infrastructure or execution, not strategy. If the answer is yes, the buyer has found a strategic partner — which is the only layer that justifies the spend.
The Coca-Cola review is the largest test case of 2026 for whether the industry has learned this distinction. The $1.7 billion is a bet on infrastructure. The strategic return will be determined by whether anyone in the operating system is accountable for the third-layer decisions the infrastructure exists to serve. The buyer has not yet answered that question. The winning bidder will not be the one with the better data infrastructure. It will be the one who can credibly answer: yes, we are accountable for the strategic decisions, and here is the unified intelligence infrastructure that makes that accountability scalable.
That is the layer an AI-native strategy agency operates in. Not a tool. Not a dashboard. Not a creative AOR. A team accountable for the strategic decisions that move your business forward — with the intelligence infrastructure to back it. One subscription. Decisions absorbed from the six to twelve tools you'd otherwise need to operate to get the same clarity. End the sprawl. Get outcomes.
Ready to end tool sprawl and get strategic accountability? Get started with Autostrat.
Book a 30-minute demo. Bring a live question and watch the answer get built.