There’s a question I’ve started asking in almost every AI strategy conversation I have with agency leaders. It’s a simple one, but it tends to make people stop in their tracks:
What is something your team does that AI can’t?
Not your tools. Not your process. Your team. Where does their presence — the judgment, the experience, the instinct about a client or a market or a piece of creative — change the outcome?
Most people pause longer than they expect to.
That pause is where the real AI conversation begins.
The dominant framing around AI in agencies right now is efficiency. Faster briefs. Faster drafts. Faster reporting. And there’s nothing wrong with that — efficiency has real value.
But efficiency is not a strategy. And if it’s the primary lens through which you’re adopting AI, you’re probably solving for the wrong thing.
Here's the problem: efficiency optimization assumes the current workflow is basically right, and that the goal is to run it faster. But a lot of agency workflows aren't basically right. They're the accumulated result of decades of habit, client expectation, and structural inertia. Making them faster doesn't make them better — it just makes them faster.
There's something we say in our AI workshops that cuts even deeper: AI is an amplifier. If you're really good at something, AI will make you even better at it. But if there are cracks in your foundation — your workflows, your leadership alignment, your billing model — AI will amplify those too. The tool doesn't discriminate. It just accelerates whatever is already there.
So the better question isn't how do we do this faster? It's why are we doing this at all, and where does human judgment actually move the needle?
I’ve started using the term “high-leverage moments” to describe the places in a workflow where a human — specifically a skilled, experienced, judgment-driven human — changes the outcome in ways AI can’t yet replicate. Not because AI is bad, but because some things require something AI doesn’t have.
What does that look like in practice?
It looks like a creative director who hears a client presentation and immediately senses the unspoken tension between what the marketing team wants and what the CEO will actually approve — and knows how to navigate it. It looks like a strategist who has been in the room with a brand for years and understands the emotional third-rail topics that never appear in a brief. It looks like a senior account leader who knows that a client’s silence on a draft isn’t approval — it’s anxiety — and calls them instead of waiting.
These aren’t soft skills. They’re hard-won capabilities that exist at the intersection of domain expertise, relationship context, and human judgment. They are genuinely irreplaceable. And right now, they’re buried.
They’re buried under status updates, formatting, routing, summarizing, transcribing, scheduling, and a hundred other tasks that are not low-importance — they’re just not human-important. They require competence, not judgment. And competence, increasingly, is something AI does exceptionally well.
When I work with agency leaders on AI transformation, I push them toward a specific reorientation: instead of starting with what can AI do for us, start with where do humans matter most, and then build backward from there.
This shifts everything. You’re no longer asking AI to speed up your current process. You’re asking AI to clear the runway so your best people can show up where they actually change things.
It also forces a harder conversation. Because when you map out where human judgment genuinely changes outcomes — and where it doesn’t — you often discover that some of the work you’ve been treating as high-value isn’t. Not because your people aren’t talented, but because the work itself doesn’t require their best capabilities. It requires competent execution. Which is different.
That realization is uncomfortable. It can feel like a threat rather than a clarification. But agencies that get through that discomfort — and redesign their workflows around where humans actually matter — come out the other side with something more valuable than efficiency. They come out with a clearer story about what they’re actually selling.
There’s a framing I keep coming back to when thinking about AI strategy in agencies. It goes something like this: Are we making humans more important in this process, or less?
Not fewer humans. More important humans. Humans whose presence at the right moments carries more weight, not less, because everything routine has been handled.
That’s the version of AI adoption worth pursuing. Not the one where everyone is scrambling to use AI before someone notices they didn’t. The one where the senior strategist is freed from an inbox full of status updates and spends that time having the conversations that used to require three follow-up emails to schedule.
The agencies that figure this out won’t just be more efficient. They’ll be more valuable — to their clients, to their people, and to themselves.
The runway matters. But it only matters because of what takes off from it.