AI won't replace your marketing team. It'll replace the one doing the wrong things.
AI is compressing the execution layer. Strategy, judgment and commercial accountability are becoming rarer and more valuable. Where does that leave your agency or in-house team?
What AI is actually doing to the work
The adoption figures are real: 78 percent of organisations now use AI in at least one function, with 88 percent reporting regular use. But look one layer down and the picture changes. Only about 21 percent have redesigned their workflows, around 39 percent report measurable EBIT impact, and just 5.5 percent qualify as AI high performers. Most organisations have adopted the tool without changing the work, which is why the promised productivity has not landed for most of them.
Where it has landed, it compressed execution. Klarna cut its image production cycle from six weeks to seven days and roughly halved its marketing team. The execution layer, the repetitive production of assets, is exactly what AI does well, and exactly the layer that is being squeezed.
The uncomfortable part: it's mostly executives, not AI
Here is the finding the headlines skip. Around 60 percent of companies cut headcount anticipating AI value, against about 2 percent reporting cuts from realised value, a thirty-fold gap. The layoffs are running ahead of the results. AI is frequently the stated reason for a decision that was going to be made anyway.
And the most aggressive example is instructive. Klarna, having automated hard and cut deep, began rehiring in 2025, citing quality. The lesson is not that AI cannot do the work. It is that the wrong work got automated, and judgment turned out to be the thing that mattered.
What gets rarer, and therefore more valuable
As execution compresses, three things become scarce and prized: strategy, judgment and commercial accountability. The person who can decide what to make, tell whether the AI output is any good, and stand behind the number is worth more, not less. The person whose entire role was producing the assets is the one exposed.
This is already visible in the agency market, where roughly 15 percent of US roles were eliminated in 2026 after about 8 percent the year before. The roles going are executional. The roles growing own outcomes. The response is not to fear AI, it is to move up the value chain toward the work it cannot do.
AI isn't taking marketing jobs, executives are, and using AI as the alibi. Thirty times more firms cut on anticipated value than realised value, and the one company that automated hardest is rehiring. The wrong things got automated.
Key research
- 78 percent of organisations use AI in at least one function; only about 21 percent have redesigned workflows; around 39 percent report measurable EBIT impact; only 5.5 percent are AI high performers. McKinsey, State of AI 2025
- Klarna cut its image cycle from six weeks to seven days and roughly halved its marketing team, then began rehiring in 2025 citing quality. Klarna; Fortune, May 2025
- Around 15 percent of US agency roles were eliminated in 2026, after about 8 percent in 2025. Forrester Predictions 2026
- Roughly 60 percent of companies cut headcount anticipating AI value versus about 2 percent from realised value, a thirty-fold gap. Davenport and Srinivasan, January 2026
- Agentic marketing is shipping: Salesforce Marketing Cloud Next and Agentforce, October 2025. Salesforce
Questions senior buyers ask
Will AI replace my marketing team?
It will replace the parts of the work that are pure execution, and it will make strategy, judgment and accountability more valuable. The exposed role is the one whose entire job was producing assets. The protected role is the one who decides what to make and stands behind the result.
Why hasn't AI delivered the productivity we were promised?
Because most organisations adopted the tool without changing the work. About 78 percent use AI but only around 21 percent have redesigned their workflows, and just 5.5 percent are high performers. The value is in the workflow redesign, not the tool licence.
Is it true companies are cutting jobs because of AI?
Often it is the stated reason more than the real one. Around 60 percent of companies cut headcount anticipating AI value against about 2 percent reporting cuts from realised value. The layoffs are running ahead of the results, with AI as the convenient explanation.
What should our team focus on to stay valuable?
Move up the value chain toward what AI cannot do: deciding strategy, judging quality, and owning the commercial outcome. Let AI take the repetitive production. The Klarna reversal, automating hard then rehiring for quality, is the cautionary tale.