Bringing AI into a creative agency without breaking it.

The people who run creative agencies already know AI is fast. Raise it at the next leadership meeting and watch the temperature change. Three things sit under the surface. Get them right and AI becomes a real advantage. Ignore them and you'll spend a year fighting your own team.

The pricing problem you can't avoid

The maths is uncomfortable. Your agency probably sells time. You bill by the hour, the day, or the headcount on a retainer.

Now picture a concept that used to take an art director and a copywriter three days. With the right AI workflow, it takes an afternoon. You've just done the work in a fraction of the time, which means you've billed a fraction of the fee.

An agency built on selling hours has a real problem here. The better you get at AI, the less you earn for the same output. That's not a reason to avoid it. It's a reason to change how you charge.

The agencies that come out ahead move away from selling hours and toward selling outcomes. Value-based pricing. Productised services with a fixed fee. Retainers tied to growth rather than timesheets.

There's also a cleaner way to frame AI for your team. Point it at the unbillable admin that already eats your margin. Use it to protect the profit on every job and free up senior people for the strategic work clients will pay a premium for.

The bit about creative pride

This one's human, and it's the part most consultants skip.

Designers, writers and directors don't just make the work. They are the work. Their identity is wrapped up in what comes out the other end.

So when a machine produces a passable mood board in four seconds, something in the room tightens. Tell that person AI is "just a tool to make you faster" and they'll see straight through it. The pushback won't be loud. It'll be quiet: tools left unopened, workflows made more complicated than the job actually needs.

You can't manage that with a memo. You manage it by redefining where a creative's value sits.

AI collapses the time spent on execution, the mechanical part of the job. The value moves upstream, into taste and judgement. The creative who thrives sets the direction, stays in the loop, and decides what's good enough to leave the building.

Call it the human-in-the-loop. The machine generates the options. The person applies the judgement that stops mediocre work reaching the client. That's a more senior and more interesting job than most creatives have now, and it's worth saying so out loud.

Fear, secrecy and the slop that follows

When people think AI means layoffs, they hide it. Your best people will use it on the quiet, cover their tracks, and never mention the tool that helped them finish early. Nobody volunteers a time-saver when the reward for saving time is a smaller team.

That secrecy has a cost, and it's an ironic one. Frightened, unguided AI use produces the thing that damages your reputation: generic, lifeless concepts that need hours of human rescue before any client sees them. The team hides the tool, the tool churns out filler, and the filler lands on your desk as a problem to fix.

Cutting headcount the moment AI arrives looks like a clean saving on paper. In practice it strips out the one thing clients pay you for: a point of view they can't get anywhere else. You bank a small cost reduction and hand away your real advantage.

The fix is the opposite of secrecy. Make it safe to experiment, and say so plainly. Set clear guardrails for what people can try and where the lines are. Then point the whole team at one number: revenue per employee.

That number changes the conversation. "How few people can we run on" becomes "how much more can this team create for each client." More variants tested against each brief, a wider creative range on offer, without burning anyone out or hiring ahead of the work.

Where to start

You won't fix pricing, culture and fear in a single workshop. Start by being honest about which one is really holding your agency back.

For most agencies, the technology isn't the blocker. The tools are ready. The harder questions are commercial and human, and they're the ones worth answering first.