Senior Marketers Are Back to Doing the Work. That’s Not a Problem.

Stop. Read This First.

Before you read anything else this week, go here: A New Year’s Letter to a Young Person by Luis Garicano. It will take you twelve minutes. Forward it to your team when you’re done.

Garicano’s argument, stripped down: AI is very good at single tasks. It is terrible at messy ones. And the messier your job, the more valuable you become, provided you pick up the tools.

We talk to senior marketers every day. Here’s what we’re hearing.

You’re a doer again.

Something shifted. It happened quietly, somewhere between the reorg and the agency review and the budget cut that forced you to just handle it yourself. The career arc that once moved from doing to directing has reversed. You are back at the keyboard. You are writing the brief, pulling the data, building the deck.

For a lot of marketers, this feels like a demotion. It is the opposite.

Garicano’s insight is that messy, multi-threaded work is exactly what AI cannot automate. Running a brand that spans retail, e-commerce, and shopper marketing, figuring out left pocket or right pocket on a Walmart investment in real time, managing the politics of a Marrtech consolidation while keeping a 30-year veteran on the train. That is not a single task. That is the bundle. And the bundle, done by someone with 20 years of judgment, is irreplaceable.

One marketer we spoke to recently described her role this way: she oversees strategy, she sits in customer meetings, she does some of the day-to-day on the biggest accounts. Her team is three people. She has to be scrappy in a way her title suggests she should not have to be. She is not complaining about it. She is winning with it.

The tools change everything, if you actually use them.

Here is the uncomfortable part. The marketers who are pulling ahead are not the ones with the biggest teams or the most sophisticated tech stacks. They are the ones who picked up Claude or ChatGPT six months ago, got over the awkward phase of using it like a search engine, and started building with it.

And here is the tension we hear constantly: the companies most capable of adopting these tools are often the ones with the most guardrails against using them. One VP told us her entire stack is locked down to Copilot on the web. She knows exactly what the stakes are, not just for her performance today but for her relevance tomorrow. If the culture does not catch up, she will leave for one that has. Another VP, at a major beverage company, took a different approach. His company told him not to bother. So he built on his own time, brought the work back in, and walked senior leadership through what he had made, and how. Minds were blown. He did not ask permission. He made the case. The question of where to invest your next few years is increasingly inseparable from the question of whether your organization is serious about this shift. Loyalty to a company that is not is a career risk. But you can also be the one who changes the company’s mind.

One more thing about the tools that almost nobody says out loud: they do not just speed up the work you were already doing. They open paths you never would have gotten to because you could not get out of the first one fast enough. A colleague described it recently as suddenly feeling overwhelmed by ideas, not by work. For the first time in her career, her thinking was outrunning her capacity to execute, instead of the other way around. That is a completely different problem to have. It is a good one.

The real unlock is your team.

You have spent a career doing messy work. Reading the room. Navigating the matrix. Making the call when the data is incomplete. That is not intuition, that is accumulated judgment. And it is exactly what Garicano says AI cannot replicate.

The leaders who are building something exponential are the ones teaching their teams to do what they know how to do. A senior marketer at a global hotel company has been running what she calls a transformation for three years. She brings in her biggest skeptics on the pilot teams, specifically. Let them tear it apart. What are we not thinking of? What would spike your anxiety? Her most protective overthinkers have become her loudest advocates. She told us: if you are brand new to AI and we are all learning together, this is your single best moment to shape it.

Garicano writes that AI gives individuals the leverage that once required an organization. That math is available to marketing teams too, if the leader has the skills and the team has the training. The long division, as one marketer put it, is what lets you write the right prompt. Your years of it are the asset. The question is whether you are deploying it.

What to do with all of this.

1. Get personally efficient, now. Not eventually. Are you actually building in these tools or just asking them questions? Have you used Claude’s desktop tool yet? The gap between people using AI as a chat interface and people building with it is growing every month.
   
   
2. Audit your company’s posture. Is it serious about this, or performing seriousness? Copilot as the only sanctioned tool is not a strategy. You owe it to yourself to know the answer, because your next career move may depend on it.
   
   
3. Model the behavior for your team. Walk them through the prompt, the thread, the output, the edit. Most people need to see it once before the instinct kicks in.
   
   
4. Redefine what your team is for. Not executional output. Judgment. The capacity to do the messy, contextual work that your brand actually needs. That is how you extend your own expertise across a broader surface area, and it is an exponential return on everything you have spent the last two decades building.
   
   

The ground is shifting. You already know how to work on shifting ground. The only question is whether you pick up the tools fast enough to make that an advantage instead of a liability.