Building a Creative Services function from zero.
A new role, a new function, and a department that had never been measured. I built the operating layer, then the numbers to show whether it was working.
- 11consecutive profitable months
- ~20%average Creative gross margin
- ~39%peak Creative gross margin
- ClickUpagency-wide rollout
The apparent problem
Hype is a fully remote agency. When I joined as Head of Creative Services, the role was new and the function did not exist yet. At the busiest point the work covered around 25 core creatives plus a freelance pool, 20-plus live client accounts, and six to eight Account Directors, spread across a lot of timezones.
Work was planned across whichever combination of Slack, Notion, Sheets and ClickUp a given team preferred. Creative had no departmental P&L, so nobody could say what the department cost or earned. The weekly meeting was doing a lot of heavy lifting.
Make the work visible
Make the work visible enough to plan, cost and measure. Then run the department on that, rather than on memory.
Build control
I ran the existing system for a few weeks before replacing any of it, because the declared process and the real one are rarely the same document.
Then I moved Creative properly into ClickUp and brought Account leads into the same view. The software was not the intervention. Having one place where work existed in a form other people could inspect was: who owned it, what had been briefed, when it was needed, who was allocated, what had changed.
On top of that I built the parts a department needs to be operable. Intake, so poorly formed work could be questioned before it reached the team. A forward capacity view, so Account teams could see when work could realistically start. Freelance forecasting, so external spend stopped being a series of surprises. And a departmental gross-margin view comparing allocated revenue against permanent staff and forecast freelance cost.
I also ran workshops with Account teams on briefing, creative project management and the client-facing habits that decide how delivery actually goes.
Commercial proof
The Creative P&L started negative. It moved into 11 consecutive profitable months at approximately 20% average gross margin, peaking near 39%. Not perfect accounting to the last subscription, but enough instrumentation to make the conversation about mechanism rather than impression.
ClickUp went from partial use to agency-wide adoption inside a quarter.
My manager rated every review category 5 / 5, Exceptional:
"barney has almost single handedly built the csm function from 0 to what it is now."
Manager review
The same review described me as a "systems thinker and excellent communicator", and recommended taking the approach into client experience, People, Sales and Strategy. That is roughly what the second half of the role became.
The boundary moves upstream
Three things I took out of it.
Instrumentation changes the argument. When work, cost and capacity are invisible, organisations end up debating impressions. When they are visible, you can debate the mechanism instead, which is a much shorter conversation.
Changing the shape of a team is not the same as changing how it operates. A new structure can redistribute an identical problem very tidily.
And a correct diagnosis creates no authority whatsoever. Change needs sponsorship, sequencing and enough organisational capacity to hold it. Most of the second half of the role was that, not analysis.
The end is worth saying plainly. Creative client revenue contracted sharply and delivery capacity became the priority. My role was largely non-billable, so I stepped back rather than protect management overhead while designers were at risk. The wider work was unfinished when I left. That is the boundary of the case, not a footnote to it.
One department, read eight ways.
The same system each time. What changes is where the boundary is drawn, and therefore what counts as the problem. Creative is in the same place in every frame.
Creative is the problem.
That was the brief, and at the start the symptoms supported it. Overworked team, reactive resourcing, repeated rounds of iteration. One department, clearly at fault.
Everything arrives here at once.
Late requests, weak briefs, promises already made to a client, scope that moved after the fact. Five upstream functions, one place where it all becomes someone's afternoon.
Nobody is looking at the same thing.
Slack, Notion, Sheets, Docs, calls, DMs and memory. Even where the same tool was in use, there was no consistent way of representing the work. A weekly meeting reconstructed reality from scratch.
One shared representation of reality.
The same information, in one place, in a form other people could inspect. Who owned it, what had been briefed, when it was needed, who was allocated, what was late, what had changed.
Instrumentation, not supervision.
Intake could be challenged, capacity could be seen, freelance cost could be forecast, output could be traced against what had been sold. The department stopped being a black box.
The margin becomes evidence.
Negative to eleven consecutive profitable months, averaging approximately 20% gross margin with a peak near 39%. Not perfect accounting. Enough instrumentation to change what the argument was about.
The cost is created upstream and lands here.
The same nodes, read as a chain. A price set in Sales, scope absorbed in Account, a brief that arrives incomplete, and a fulfilment cost that shows up in the one department with a measurable P&L.
The boundary was drawn in the wrong place.
The more clearly the work was measured, the wider the true problem boundary became. Creative is still exactly where it was. It is no longer the subject of the diagram.