The following is a guest piece written by Natalie Bastian, CMO at InMarket. Opinions are the author’s own.
At last month’s Cannes Lions International Festival of Creativity, I sat down with CMOs of major global brands to discuss how artificial intelligence is reshaping advertising. Going in, I expected to hear excitement about the wins. Instead, many conversations turned to accountability and how far behind advertising is falling.
AI is maturing fast. The strategies are real. But most organizations still don’t have the muscle to act on what AI tells them, and you can’t build muscle without the right inputs. Good data is the protein. Without it, you’re just going through the motions. That gap between what AI makes possible and what companies actually do with it was the most important thing I took away from Cannes this year.
A new AI reckoning
Last year at Cannes, people were still asking what to do with AI: running pilots, exploring internal use cases, figuring out how to position it to clients. This year, the question changed.
“Pilots are over, everyone is in production phase,” Sanjna Parulekar, senior vice president of product marketing at Salesforce, said during a panel at Cannes.
From my perspective, that’s exactly the shift underway. AI is no longer about experimentation for experimentation’s sake: It’s about building the accountability, orchestration and measurable outcomes that turn intelligence into business impact. Amazon Ads has built a conversational planning interface that takes a marketer from insight to activated campaign without a single SQL query. PMG has repositioned itself from a media services company into a technology and transformation company.
But production doesn’t mean progress. Most brands still have more data than they know what to do with, and they’re still optimizing channel by channel instead of designing for outcomes across the business.
Naming the accountability gap
According to Forrester data published in 2024, 64% of B2B marketing leaders don’t trust their organization’s measurement for decision-making. When the data isn’t trusted, the intelligence built on top of it won’t be trusted either. That’s why I keep coming back to a simple truth: You can’t blame AI for the output. It’s only as good as the question you ask. AI doesn’t eliminate the accountability gap: It makes it more visible. If we don’t have clear ownership of the data, the inputs and the outcomes, AI will only accelerate the production of answers no one fully trusts.
AI can drive efficiency, whether in media buying or audience targeting, but it also raises the stakes. The real risk is not just bad automation. It’s scaled automation built on weak inputs, unclear accountability and measurement that still can’t prove business value.
Causation is the new currency
Speed driven by AI only becomes an advantage when you can prove what it produced. CFOs are asking marketing the same question they ask every other function: What did this cause? Most of the CMOs I spoke with at Cannes don’t have a definitive answer.
For years, impressions, reach and clicks were the currency because those numbers were easy to produce and easy to defend. Proving causation was slow and expensive, so correlation became the default. That excuse doesn’t hold anymore. The brands investing in causal measurement now are the ones building a competitive moat.
“The goal is to position marketing as a growth driver, not a cost center,” Julia Fedor, head of brand marketing operations at United Airlines, told me.
That’s the bar marketing needs to clear now, and it takes metrics built for causation, not correlation.
What CMOs can do about it
The gap isn’t just a technology problem. Here’s where to start treating it like the organizational one it actually is to maximize your AI investments:
- Retire one vanity metric this quarter. Pick the KPI that your team tracks, but nobody acts on. Removing it signals that the accountability standard has changed.
- Audit what you act on. Go back through your last several quarterly business reviews and identify which metrics drove a decision that changed something or proved impact. If you can’t find one, fix that before you buy more tooling.
- Get your CFO in the room earlier. The CMOs building the most organizational trust don’t wait for budget season to prove marketing’s value. Bring the CFO into the conversation about what success looks like early on, before a campaign even runs.
- Separate the AI roadmap from the decisioning roadmap. Most organizations are investing heavily in AI, but not in the change management that surrounds it. The technology can accelerate the output, but it needs to be used correctly to drive outcomes.
- Define what “growth driver” means in your business. AI can only drive growth when the definition is concrete. Pick the metric that moves the needle — incremental sales lift, revenue contribution, new customer acquisition — and use it as your proof.
What Cannes got right
The best conversations I left Cannes with weren’t about new AI use cases. They were about what it takes to build an organization capable of acting on that intelligence.
The brands winning right now are the ones with the clearest line between what they learn and what they do. They’ve decided not only that AI will drive their business forward, but also, more importantly, that they’ve built the structure to convert intelligence into measurable outcomes.
That’s how you cross the threshold into the era defined by outcomes and close the gap between data and decision. Speed gets you to the table. Proof is what keeps you there.
