There is a version of the future that agencies are quietly building for themselves where strategy is fast, scalable, and hollow.
It is produced by AI. Organised by frameworks. Validated by data. And it is indistinguishable from every other agency’s output, because it is built from the same corpus, trained on the same internet, optimised for the same legibility.
I want to talk about what gets lost in that version. And why the cost of losing it is far higher than any headcount saving will account for.

Data tells you what happened. It doesn’t tell you what it means.
I’ve been thinking about this for two years now. Not abstractly — practically, in the room, on live briefs.
Data can tell you that a behaviour occurred. Frameworks can organise what you already know. AI can synthesise research at a speed no human can match. None of that is trivial. All of that has a place.
But the ability to stand inside a cultural moment — to feel its texture, understand what it is actually asking of brands, name the thing beneath the thing — that requires being a human being who pays attention. A specific kind of human being. One who is currently living inside the culture they are tasked with interpreting.
That is not a soft skill. That is the skill. The one that determines whether a strategy unlocks something real or produces something correct-but-inert.
The junior planner is not a cost. They are your primary research instrument.
Here is the argument agencies are not making loudly enough:
Junior and mid-level planners are not apprentices waiting to become useful. They are the people closest to the lived experience of the audiences most brands are trying to reach. They are inside the subcultures. They are fluent in the codes. They are the ones who can tell you — not from a trend report, but from a Tuesday — what something actually feels like on the ground.
When you cut that layer, you do not just lose capacity. You lose signal.
You lose the person who can tell you that the cultural reference in the brief reads as three years out of date. Who can say, without a deck to support it, why the tone feels wrong for this moment. Who brings the consumer’s actual voice into the room because they are part of that consumer’s world.
Frameworks don’t do that. Syndicated research doesn’t do that. AI, trained on what has already happened, certainly doesn’t do that.
If everyone uses the same tools, everyone produces the same work.
This is the structural risk that no one is talking about directly enough.
When agencies rely on AI for research and insight generation, they are all drawing from the same well. The same training data. The same synthesis patterns. The same tendency to produce the confident, well-organised, and entirely predictable.
The result is not efficiency. It is convergence. A category where every strategy presentation sounds like it was written by the same author, because in a meaningful sense, it was.
The differentiator — the thing that allows one agency’s work to feel genuinely surprising, to land with a kind of accuracy that the audience didn’t know they needed — is human specificity. The planner who has a perspective that was not generated. Who brings a point of view that cannot be aggregated from existing data because it is emerging in real time, through the experience of being a person in the world.
Cut that layer, and you are not just slimmer. You are more similar to everyone else. And in a commoditised market, similarity is the most expensive position of all.
The culture of an agency is also the quality of its work.
There is one more argument that doesn’t get made in the cost-cutting conversation, because it is harder to put in a spreadsheet.
Junior and mid-level planners are not only producing insight. They are generating the intellectual energy of an agency. The questions that challenge a received idea. The cultural reference that re-frames the brief. The willingness to say this doesn’t feel right before the work has calcified into a recommendation.
That energy is generative. It is what keeps senior strategists from operating in their own echo chambers. It is what makes creative teams feel like they are being briefed by someone who actually knows something, rather than someone who has organised what is already known.
The agencies that will continue to do groundbreaking work are the ones that do not mistake that energy for overhead. It is not overhead. It is the engine.
The brief I’m making.
Cost-cutting in planning is a rational-looking decision that is strategically irrational.
You save on salary. You lose on signal, originality, cultural fluency, and the capacity to produce work that is genuinely differentiated.
You build a team that is faster, cheaper, and less able to do the thing that justifies the premium you charge.
Cultural fluency is the last advantage that cannot be automated. Not because AI won’t try — it will — but because it requires being inside a culture in real time, with a body and a perspective and an attention that is not retrospective.
That is what junior and mid-level planners offer. Not potential. Not talent in waiting.
Contribution. Now. That cannot be replicated.
The question for any Head of Strategy is not whether they can afford to keep that layer.
It is whether they can afford to lose it.
This is the argument I keep coming back to. The one I think matters most right now. I’ll keep working it out here, in public — because the thinking is better when it has to hold up to scrutiny.



Brilliant article. This is also the reason tools like Canvas8 need to be invested in rather than the standard ‘data tools’ . Like the junior strategist, a source of context to bring the why behind the data into agencies