Who Trains Your Next Senior Trader?

The biggest risk AI poses to AdTech and media teams may not be mass job losses.

It may be something quieter.

The repetitive work that used to teach junior traders how campaigns actually behave is increasingly being automated. And if teams remove the work without replacing the learning, they may discover in three or four years that they have plenty of tools — but far fewer people ready to become senior.

The teams that redesign junior roles now won't just become more efficient. They'll build the experienced talent everyone else will eventually be trying to hire.

Key takeaways

• AI is removing some of the repetitive execution work junior people traditionally learned through.
• That work wasn't always exciting, but it gave people repetition, pattern recognition and exposure to mistakes.
• Removing tasks is not the same as removing the need to learn what those tasks taught.
• Junior roles need to shift towards supervised judgement, diagnosis, QA and decision-making.
• Managers will need to make more of their own thinking visible instead of assuming experience will accumulate naturally.
• The question for leaders isn't simply “What can AI do?” It's “How will someone learn this once AI does it?”

The apprenticeship nobody called an apprenticeship

Think about how an experienced trader became experienced.

It probably wasn't through one training programme.

They built campaigns.

They checked pacing.

They changed bids.

They pulled reports.

They investigated discrepancies.

They watched a campaign underperform and had to work out why.

They sat beside someone more experienced and heard why one recommendation made sense while another didn't.

They made small mistakes when the stakes were relatively low.

And then they did it again.

A lot of early-career work is repetitive precisely because repetition is part of how expertise develops.

Nobody necessarily called that an apprenticeship.

But functionally, that's what it was.

The junior did the work. A manager reviewed it. Over time, the junior started seeing things before the manager needed to point them out.

That's how judgement gets built.

AI changes more than the workload

The obvious benefit of automation is that teams no longer need people spending hours doing work technology can complete in minutes.

That's good.

The problem comes when businesses treat the disappearing task and the disappearing learning opportunity as the same thing.

Gartner warned in September 2026 that as AI automates foundational marketing activities, organisations risk creating a leadership-pipeline gap because fewer employees get the experience needed to develop judgement, business acumen and contextual expertise. Its recommendation is not to preserve low-value work indefinitely, but to redesign junior roles so people are exposed to higher-value decisions earlier.

That distinction matters.

You don't need a junior trader manually doing something inefficient simply because their manager did it ten years ago.

But you do need to ask:

What were they learning while they did it?

If campaign setup becomes largely automated, where does the junior learn how a badly structured campaign looks?

If optimisation recommendations are generated automatically, where do they learn when not to follow one?

If reporting is produced in seconds, where do they learn to distinguish a real insight from an interesting-looking fluctuation?

The task can disappear.

The learning objective can't.

What a junior trader needs to learn hasn't disappeared

The tools are changing quickly.

The underlying questions are more familiar.

A trader still needs to understand:

• Why is this campaign performing differently from expectation?
• Is the data trustworthy?
• What is driving the result?
• What should change next?
• What happens if we make that change?
• Is the platform optimising towards the same outcome the client actually cares about?
• Is an unusual result an opportunity, an error or noise?
• How do I explain this clearly to someone who doesn't spend all day inside the platform?

AI can help answer some of those questions.

But eventually somebody has to decide whether the answer makes sense.

That's the skill worth protecting.

Junior roles need more diagnosis, not just less execution

One response to automation is simply to shrink the junior role.

Fewer people. Less execution. Senior traders use AI to move faster.

That may work operationally in the short term.

It creates a harder question later.

Where does your next senior trader come from?

Gartner's broader 2026 research found that 22% of CHROs said at least one business leader in their organisation had stopped hiring for entry-level roles because of AI automation. Gartner cautioned that eliminating those pipelines can force companies to compete later for experienced external talent while reducing the low-risk opportunities through which people traditionally developed judgement and institutional knowledge.

For AdTech teams, the alternative is to redesign junior work rather than simply remove it.

A junior might spend less time manually executing routine optimisation.

But more time could go into:

Auditing.
Give them an automated recommendation and ask whether they agree.

Diagnosing.
Show them an underperforming campaign and ask for three possible explanations before showing them what actually happened.

QA.
Make them responsible for checking automated setup, targeting, measurement and reporting.

Explaining.
Ask them to turn platform output into a recommendation someone outside the trading team can understand.

Comparing.
Show them two possible actions and ask what each would optimise for — and what each might sacrifice.

The objective isn't to make junior work artificially difficult.

It's to move the learning closer to the judgement they'll eventually be hired for.

Make senior thinking visible

There is another part of the old apprenticeship that's easy to lose.

Proximity.

A junior didn't only learn from completing tasks.

They learned from seeing how experienced people thought.

The senior trader looking at a strange result and saying:

“I'm not worried about that yet because…”

The manager rejecting the platform recommendation because:

“That would improve the metric, but not the client's actual objective.”

The person spotting a tracking issue because:

“That number doesn't behave like that normally.”

Those small explanations transfer context.

If AI handles more of the process privately and senior people simply approve the output, juniors may see the final decision without seeing the reasoning behind it.

Managers need to become more deliberate about narrating decisions.

That could be as simple as building five minutes into a review to ask:

What did you notice?

What would you do?

What information would change your mind?

Then explain the senior view.

The point isn't for the junior to guess correctly every time.

The point is to give them repeated exposure to how experienced judgement works.

Give juniors decisions they can safely make

You can't develop judgement without making decisions.

But handing a new trader a major account and hoping they learn quickly isn't a development strategy either.

Teams need a better middle ground.

Give junior people decisions with boundaries.

For example:

They can recommend the optimisation, but a senior approves it.

They can make changes within agreed spend thresholds.

They own QA for a particular part of the campaign.

They investigate the anomaly and present the likely cause.

They draft the client recommendation before seeing what the senior would suggest.

As capability grows, expand the boundary.

This creates what traditional execution used to provide naturally: repeated practice with feedback and gradually increasing responsibility.

Gartner similarly recommends giving early-career employees higher-value responsibilities earlier while surrounding them with coaching, guidance and other “development safety nets.”

Don't make “good with AI” the new entry requirement

There's an obvious temptation here.

If AI handles more execution, simply hire juniors who already demonstrate unusually strong judgement.

But judgement usually comes from context.

Expecting someone at the start of their career to think like a senior before anyone has given them the experiences that build senior thinking creates a circular problem.

You need experience to get the job.

But the jobs where people used to get that experience are disappearing.

Instead, junior hiring may need to look harder at the ingredients from which judgement develops:

Curiosity.

Attention to detail.

Comfort asking why.

Ability to explain reasoning.

Willingness to challenge an answer respectfully.

Ability to learn from feedback.

Strong fundamentals.

You are not hiring a finished trader.

You're hiring someone capable of becoming one.

Audit the apprenticeship before you automate it

Before automating another part of a junior role, ask two separate questions.

The first is operational:

Can technology do this work faster or better?

The second is developmental:

What did doing this work teach someone?

If the answer to the second question matters, decide where that learning will happen instead.

It might move into simulations.

Shadowing.

Campaign reviews.

QA ownership.

Structured diagnosis exercises.

Weekly case discussions.

Or real decisions with controlled risk.

The format matters less than the principle.

Every learning opportunity removed by automation needs an intentional replacement.

Otherwise the cost may not appear on this quarter's headcount plan.

It appears years later when you're trying to hire someone with five years' experience that nobody gave them the chance to build.

Three questions for your team

As you review roles for 2027, ask:

Does this role still need to exist in its current form?

Which responsibilities should disappear, change or move as AI becomes part of the workflow?

Which parts of the work require human judgement?

And how will someone early in their career get enough practice making those judgements?

Could some of the execution be managed from another market?

If so, what capability, context and management structure needs to remain around it?

The answer may look different across every team.

But the organisations thinking about these questions now are more likely to have the senior capability they need later.

Who trains your next senior trader?

AI can shorten the path between a question and an answer.

It can't automatically create the experience required to know whether that answer is right.

That's still a people-development problem.

The opportunity for AdTech teams isn't to preserve yesterday's junior job.

It's to design a better apprenticeship for tomorrow's senior one.

‍

We're speaking to AdTech and media leaders about how AI is changing their teams for our H2 Talent Report.

If you're rethinking junior roles, we'd love to hear how.

chat@uncommonlygoodpeople.com

And if you're reviewing your team for 2027, start with these three questions:

Does this role still need to exist in its current form?
Which parts of the work require human judgement?
Could some of the execution be managed from another market?