Marketing has spent the past few years asking how AI can make existing work faster. How many briefs can we produce, campaigns can we launch, reports can we automate and customer interactions can we personalise? But I think we are asking the wrong question. The bigger opportunity is to ask whether we should still be doing the work in the same way at all.
Over the past few years, I have been leading that transformation inside my own global marketing and communications organisation. Like many companies beginning this journey, our first instinct was to chase productivity. We gave talented people new AI tools, encouraged them to experiment and quickly saw individual tasks become faster.
Research that once consumed days could be completed in hours. Content production acceleratedand analysis that required significant manual effort became far more manageable. Those gains were real. But something more fundamental hadn’t changed. Campaigns were still moving through the same approval chains, decisions were slowing at the same points and teams were still waiting on each other. We had made people faster without making the system any better.
That changed the question for me. Instead of asking how AI could make our existing marketing organisation more productive, we began asking: if we were designing the organisation today, knowing what AI can now do, would we design the work this way at all?
The answer was no. And that was where the real transformation began.
This is where many organisations get stuck. Brand teams explore AI for brand activities. Content teams automate content creation. Communications teams find efficiencies in communications workflows. Each function builds its own use cases and reports its own productivity gains. But customers experience journeys, not departments. They don’t care where internal responsibility sits. They care whether the experience is relevant, timely and connected.
When we looked closely at our own organisation, we discovered that much of the friction wasn’t inside individual roles at all. It existed between them – in the handoffs, approvals, dependencies and moments of waiting that had accumulated over time. One campaign workflow required about 80 manual tasks, 50 handoffs, 20 days and roughly 100 hours of human effort. After redesigning the work, that became 40 manual tasks, nine handoffs, three to five days and around 14 hours of human effort. The number that interested me most wasn’t the hours saved. It was the 41 handoffs that disappeared. Each represented a moment when work could stop, wait or lose context.
That experience taught us to stop starting with the organisational chart and start with the work. A job is a collection of tasks, and a workflow is a collection of tasks in sequence. Once you break marketing down that way, you can ask much better questions: what can AI lead? What should it accelerate? Where does human judgement materially improve the outcome? And which work no longer needs to exist at all? You don’t add AI to the work. You redesign the work.
Producing more content, faster, will not automatically deliver better experiences. AI can generate options, identify patterns, analyse audiences and produce enormous volumes of content. But people still determine what a brand should stand for, which idea deserves attention, how cultural context changes the message and when something simply doesn’t feel right. AI can scale content, but it cannot scale meaning. That is why the future of marketing…