AI has changed how marketing teams work, but the gains are uneven. The teams pulling ahead are not using more tools — they are using fewer, in the right places, and measuring all of it.
Where it pays off
- Creative iteration speed — testing more angles per week than a team could produce manually
- Audience and intent modelling on first-party data you already own
- Always-on optimisation across channels, adjusting between reporting cycles rather than after them
- Research and briefing — compressing the unglamorous hours before anything gets made
Where it mostly does not
Volume for its own sake. Generated content published without a point of view performs exactly as well as it deserves to, and increasingly gets filtered before a human sees it. Positioning, offer and audience remain the levers; AI moves faster along them but does not choose them for you.
The teams getting compounding returns treat AI as an accelerator on a loop that already works. The teams disappointed by it usually installed the accelerator first.