The Red Pill Series: What AI in marketing looks like when nobody’s watching

Marketing functions have the highest revenue potential from AI of any business function. They also have some of the lowest adoption confidence. In the second article of The Red Pill Series, humari’s Joe Hildebrand explores the uncomfortable gap between the AI story marketing leaders tell and what’s happening inside their teams.

According to research tracking AI adoption across the enterprise, marketing and sales functions consistently lag behind IT, procurement and finance on both the frequency of AI use and the confidence with which teams deploy it. This pattern has held since 2023. Yet data tells us that marketing and sales is the function most likely to report revenue increases from AI.

In other words, marketing sits on the single biggest commercial opportunity from AI in the entire enterprise, and it’s one of the least ready functions to capture it.

The narrative gap

That’s not the story most CMOs are telling. On conference stages and in board papers, the narrative is one of momentum. AI is a strategic priority, budgets are being redirected and productivity gains are already flowing. It sounds like a function that has this under control.

But dig beneath the surface and a different picture emerges. Only 17% of marketing professionals have received detailed AI training. And while leadership frames AI as an opportunity, the decisions being made around it tell a different story: agency relationships cut, headcount reduced, roles simplified. For the people inside marketing teams, the lived experience of AI transformation often feels less like empowerment and more like a countdown.

The super-user illusion

The problem isn’t that people aren’t using AI, in fact research has found that around 40% of marketing employees are already AI super-users. Usage, at least on the surface, is high. But it’s not translating into strategic impact. People are using AI to do the same things slightly faster, generating first drafts, summarising data, producing social copy at scale, rather than fundamentally rethinking how their work creates value. It’s surface-level adoption dressed up as transformation, and it’s exactly what the blue pill looks like from the inside. Comfortable. Familiar. And quietly going nowhere.

What Klarna learned the hard way

Perhaps nobody illustrates the danger of mistaking efficiency for transformation better than Klarna. In 2024, the fintech company became the poster child for AI-driven cost reduction, replacing roughly 700 customer service agents with an AI chatbot and cutting its workforce from over 5,500 to around 3,000. Revenue grew 108%. Every conference keynote cited it as proof that the future had arrived.

By early 2026, Klarna was quietly reversing course. Customer satisfaction had deteriorated on complex interactions and the cost of handling quality failures consumed more than the original savings. The company began rehiring, and what was once a case study in AI efficiency became a cautionary tale about what happens when you optimise for cost without investing in the human judgement that holds quality together.

What the best CMOs are doing differently

The marketing leaders who are getting this right aren’t starting with tools or adoption targets. They’re starting with honesty. They’re having direct conversations with their teams about what AI means for their roles, not just what it can do. They’re creating space to experiment without the pressure to immediately prove efficiency gains. And they’re investing in building the capabilities, including creative judgement, intuition and the confidence to challenge AI outputs, that will determine whether their function captures AI’s enormous potential or watches it erode the things that made marketing valuable in the first place.

In the next article, we’ll explore what it takes to lead through this moment. Because the gap between narrative and reality doesn’t close itself, it closes when someone decides to lead differently.