Learning AI Won't Make You Stand Out for Much Longer

Over the past two months I've been speaking with founders, CMOs, agency owners and marketing leaders about how AI is changing marketing teams. We talked about smaller teams, disappearing roles, new skills and what the marketer of the next few years is probably going to look like. One thing kept coming back that I didn't really expect. While everyone seems to be telling marketers to learn more AI, the people I spoke to weren't actually hiring based on who knew the most about AI.

Somewhere right now, a marketer is probably doing their third AI course of the year. Prompt engineering, no-code agents, maybe a certificate at the end saying AI Marketing Specialist. And I get why. Everywhere you look, the message is basically the same: learn AI or get left behind.

So I started asking a pretty simple question: what do you actually look for when hiring a marketer right now?

I expected AI skills to come up a lot. Long story short, they didn't. Not because these people don't care about AI. Pretty much all of them use it in some shape or form. But nobody started with prompts, agents or how many AI tools someone knows. So are we then focusing on the wrong thing?

We're probably overcomplicating what "AI skills" actually mean

Open LinkedIn and you see the same thing everywhere. Become an AI engineer without knowing how to code. Learn these ten prompts. Build an AI marketing team in a weekend. There are entire courses and master's programs being built around the idea that marketers urgently need to become AI experts.

Laura Mantilla Vargas, Creative Director at Nas.com, builds AI marketing agents and used to teach an AI course, so she knows that world pretty well. "You get spammed on LinkedIn with posts saying come learn all the skills, all the prompts," she told me. "Then you talk to those people and what they actually know is very little."

I see something similar from the recruitment side. A company tells us they need an "AI marketing nerd", so I ask what that person actually needs to do. "Improve timelines. Be more data driven. Automate some processes." Then you start asking more questions, and sometimes you realise they don't really know either. They know AI should probably be part of the role, but haven't figured out what that actually means yet.

I've literally had situations where it felt like someone had asked ChatGPT what AI could do for their marketing team and then started hiring for the answer. I'm not convinced that's the right way around.

So I asked the people doing the hiring

This was probably the part that surprised me most. I expected at least a few of them to say they wanted the person who was clearly ahead on AI. Instead, I got more or less the same answer every time.

Alexander Wijninga, founder and CEO of Watermelon, an AI customer service platform, had a pretty simple take: "Experience with AI is a plus. But using AI doesn't mean you apply it well. I hire people who are hungry to learn."

Aad Dekkers, an international marketing leader with nearly twenty years in B2B tech and SaaS, echoed this sentiment. AI experience helps, he said, but he'd rather hire someone curious, a good listener, unafraid to experiment and fail, using AI as a tool, not a crutch.

Mark Landman, founder of the agency DotControl and former chairman of the Dutch Digital Agencies trade body, barely looks at tools at all. "We hire on personality, not on a skill set," he told me. "We barely look at junior or senior. We look at how much complexity someone can handle."

Laura does have an AI requirement when hiring, but it's ridiculously simple: do you use AI in your own life? That's basically the baseline. From there, she's much more interested in whether someone fits the team and how they think. They can teach the tools themselves, and they've even hired people directly from hackathons rather than through a normal CV process.

Giaani Paragh, Group Marketing and Communications Director at the Normec Group, had almost the opposite concern: if you focus too heavily on skills in the hiring process, you risk completely missing the person behind them. Danielle Lambo, CEO of the HR and payroll company Kolibrie, made another point I liked: stop looking for the one person who can do everything and instead build a team where people cover each other's weaknesses.

These are very different companies, industries and types of leaders, but the answer was super consistent. They're hiring for how someone thinks, learns and handles complexity. AI is increasingly just expected.

Which makes sense when you look at the work that's disappearing

The marketing work under the most pressure right now is execution: writing another post, pulling another report, running the same channel you've been running for three years or producing twenty variations of the same asset. AI is getting very good at that work, very quickly. (Yes, I know, I keep saying this every article)

What it still doesn't do particularly well is decide what should be done in the first place. What actually matters here? Which number should I care about? Is this campaign actually bad or did we just have a bad day? Should we even be running this campaign?

That's where judgment comes in. And the more execution gets automated, the more valuable that judgment becomes.

But there is one big exception

Laura pushed back on this quite hard, because there is a technical profile she desperately wants. "Everyone learned to vibe code," she said. "That's a dime a dozen. The person I can't find is the engineer who actually understands marketing."

We've searched for almost exactly that profile before, and found around 125 people across the entire Randstad who came remotely close. So yes, there is definitely a market for people who are really technical. But I think that's very different from simply being "good at AI."

Someone who did a weekend course, knows their way around Claude and can build a basic agent is one thing. Someone who understands engineering and also understands how marketing works is something else entirely. Knowing how to use AI tools is becoming pretty normal already. Finding someone who combines serious technical knowledge with a good understanding of marketing is still very difficult.

And I think that's where a lot of the conversation around AI skills goes wrong. Basic AI knowledge is quickly becoming something companies simply expect you to have, which also means it doesn't really differentiate you anymore. You need something more. That might be technical ability, or it might be ten years of understanding a market, customer or channel well enough to know when the AI output is complete nonsense.

Pim Kos, who spent about a decade building a performance marketing agency before selling it and is now a CMO, made the point that most of the tricks are becoming commoditised because everyone has access to the same tricks now. What isn't as easy to copy is context.

If you've spent ten years inside a market, you understand things that aren't written down anywhere. You know why something that looks good on paper probably won't work, and you recognise patterns because you've seen them twenty times before. AI can give someone your playbook, but it can't suddenly give them those ten years of context. At least not yet.

Laura sees the limitations from the technical side as well. She built an AI writing tool for her own team and, in her words, it breaks all the time. An API changes, a plugin updates or credits run out and suddenly someone needs to understand what happened and fix it. Alexander had a similar experience when he put AI into a process that was already working perfectly well and ended up making the whole thing more complicated. So simply adding more AI isn't necessarily the answer either.

I was also partly wrong about specialists

In the first article in this series I argued that being too specialised was becoming risky. I still think that's true, but I probably oversimplified it.

Oscar Pakasi, a senior B2B marketing and GTM leader with a systems-engineering background, put it better: "AI makes knowledge and execution easier to reach. That doesn't turn everyone into a generalist. You still have to be deep enough somewhere to know what good looks like."

Being a specialist isn't necessarily the problem. Being a specialist who only knows how to work within one channel probably is.

If all you know is how to run Meta campaigns, and AI increasingly takes over more of the campaign work, then yes, you're at risk. But somebody still needs to look at what the AI produced and know whether it's any good. For that, you need depth.

Mark described the old model as the T-shaped marketer (yes, that one again): broad knowledge with one area of real depth. He thinks it's becoming more like a block. You're broader overall and still have real depth somewhere inside it, but mindset and adaptability become much more important than they were before.

Laura gave me a good example of this. She has an AI agent looking at Meta performance. On a bad day, the agent might flag ten different things that are supposedly going wrong. Her Head of User Acquisition looks at the exact same numbers and says no, it was just a bad day across all clients and half of this doesn't matter.

Knowing what to ignore is a skill in itself. And in this example, it's probably more valuable than knowing how to build the agent in the first place.

The other thing: you don't necessarily hire all of this

Some of it you build.

Danielle made this point when we spoke. A lot of the skills people put on their CV are trainable, while the harder part is finding the right person underneath them.

Her own marketing team went from five people to three as AI started taking over more of the busywork. The obvious move would have been to cut the youngest people first. She didn't. She kept them, put them under a senior marketer who had developed much broader experience, and started deliberately building their judgment.

I think that's an important point in this whole discussion about juniors disappearing. AI can reduce the amount of junior work, but that doesn't automatically mean you should get rid of junior people. If nobody gets the chance to build experience now, where are your seniors going to come from in five years?

So what would I actually do as a marketer?

First, I wouldn't stop learning AI. That would obviously be stupid. But I also wouldn't spend the next year collecting AI certificates and assume that's going to make me valuable.

I'd focus on getting away from pure execution and moving closer to deciding what gets done rather than just doing it. That could mean strategy, positioning, understanding the customer, running across multiple channels or managing tools and people together. Basically, get closer to the decisions that still require context and judgment.

I'd still learn the technical side of AI, but I'd go deeper than prompting. Can you connect your own data? Can you build a workflow that people actually use? Can you spot when it breaks? Can you edit the output because you actually know what good looks like? Those are much harder skills to fake than knowing which prompt structure works best this month. And if you can't do that in your current job? Do it in your daily life and save some examples for your next hiring interview.

And then I'd pick something to become very, very good at. A market, a customer type, a technical skill, a channel. Something where, after a few years, you know things that aren't sitting in a ChatGPT answer.

Because everyone is going to have access to the same tools. That's already happening. The difference is what you can actually do with them.

So where does that leave us?

I don't think the marketer of the future is the person who knows the most AI tools. Most of those tools will probably be replaced anyway.

I think it's much more likely to be the person who can learn quickly, go deep when it matters and keep changing how they work without losing the ability to judge whether the end result is actually good. That's much less exciting than "learn these 10 AI skills and future proof your career", but it's also much closer to what the people I spoke to are actually hiring for.

Learning AI in 2026 to stand out is probably a bit like learning to use Google in 2005. You absolutely need to know how to use it, but very soon nobody is going to care that you can.

So if you're doing your third AI course this year, I'm not saying stop. Just make sure there's something underneath it, because that's the part companies are actually hiring for.

If you're a marketing leader dealing with the same questions, I'd like to hear your thoughts on this. This was supposed to be the last piece in this series for now, but I have a feeling the discussion is only just getting started.