Most agency leaders can name three or four tasks that are genuinely faster now.
Brief summarisation. First-pass copy. Research pulls. Version generation. The list is real. Nobody is making it up. These are actual hours that used to take longer and now don't.
But ask a different question. Ask which line on the P&L is bigger because of it.
Most founders go quiet.
Not because the savings aren't real. They are. The problem is that between "the task got faster" and "the margin improved," something happened. The time went somewhere, and most agencies have no clear picture of where.
That gap doesn't close by itself.
There are three honest answers to where the time went, and none of them require bad intent to explain.
The first is more revisions. Output that used to take three hours carried implicit pressure toward quality — it had cost something to produce, so it had better be close. Output that takes twenty minutes is easier to treat as a starting point. Then it gets another pass. Then a third. The revision cycle expands to fill the space the production cycle vacated.
The second is more versions. When production effort drops, the instinct is to generate options rather than make decisions. The extra creative route. The backup slide deck. The just-in-case variant the client didn't ask for. Optionality feels thorough. It isn't always commercially productive.
The third is scope nobody decided to absorb. A task became faster, so the team took on more of them. Nobody signed off on the extra scope. Nobody said no either. The work expanded because the operating model didn't provide a different outcome. The space opened up, and work filled it.
Three different mechanisms. The same result: faster production, flat margin.
The tempting read is that this is a discipline problem. If people just captured the time, or held scope tighter, or made cleaner decisions about versions, the margin would follow.
That read is wrong.
The team is doing exactly what the existing operating model rewards. Finish the work, keep the client comfortable, move to the next brief. That model predates AI by many years. When AI arrived and made the team faster, the model didn't change. Only the speed changed.
And the faster a team moves inside a structure that hasn't been redesigned, the faster the time savings disappear.
This is a condition, not a bad month. It doesn't respond to working harder. You can't outrun it by adding more discipline to a model that wasn't designed to capture this kind of efficiency in the first place.
At the function level, the pattern becomes harder to ignore.
In copywriting, brief-to-draft compresses. Review and revision expand. The copywriter produced faster; the account team spent longer on passes. The net saving is less clear than it looked.
In creative direction, concept generation compresses. Client revision rounds expand, because more routes went to the client and more routes generate more opinions. The upstream got faster; the downstream got slower.
In project management, status updates and task lists compress. But AI-generated handoffs tend to look cleaner than they are, because they smooth over the caveats and the uncertainties the next person actually needs. Rework arrives later, downstream, attributed to "client changes." The PM saved an hour. The delivery team lost two.
The same shape, playing out across functions at slightly different speeds. A pattern across the whole agency, not a problem in one corner of it.
This is a visibility problem before it's a measurement problem. You cannot capture what you cannot see. And most agencies do not yet have a clear picture of where AI is actually touching the work — which functions, which workflows, which handoffs, which review stages — let alone what the time is doing once it's been saved.
82% of marketing leaders say their AI initiatives are still stuck in pilot purgatory. That isn't caution. It's the same visibility gap playing out at the strategy level instead of the workflow level.
The useful first move is not a new tool, a policy document, or a training session. It's a visibility pass. Mapping what is already happening across the functions where AI has already arrived, before anyone tries to optimise or govern or transform anything.
That's the door I'd point at: a two-week diagnostic that helps agency founders see how AI is actually being used across the team, where the workflow and review boundaries are unclear, and what needs tightening first. Not a transformation programme. Not a compliance exercise. A clear picture of where the time is actually going.
If you don't have that picture yet, it's worth asking yourself honestly:
If someone asked you today where exactly the saved time went, what would your answer be?
If that question made you pause
The gap between your team producing faster and what your P&L is telling you is exactly what the AI Workflow Clarity Audit is designed to surface.
It is a two-week diagnostic. It maps where AI is entering the work across the Core Six functions: client service, strategy, copywriting, creative, design, and project management. It identifies where workflow and review boundaries are unclear, where client-data assumptions have never been tested, and what a consistent agency-wide answer would actually need to contain.
The outcome is a clear picture and the three things worth tightening first. Not a policy document. Not a transformation programme.
It costs £500. It does not require a project team.
If the scenario at the top of this newsletter felt familiar, it is probably worth a conversation.