Changing how an agency works is like changing a tyre on a moving car. The jobs are live, the clients are waiting and the team is already full, but the change still has to happen.
Every agency owner I speak to wants to improve something. Better handovers, a proper timesheet habit, a tighter briefing process, a planning tool people actually use. The ideas are rarely the problem. What gets in the way is that nobody can stop the work to make the change, so the old way and the new way have to run together for a while.
Why good ideas lose momentum
When a change fails in an agency, it rarely fails with a bang. Nobody stands up and says the new process isn't working. It just gets pushed back a week, then another week, until nobody mentions it any more.
The reason is simple. The old way is always easier in the moment. When a deadline is looming, people go back to what they know. The new process loses to a client call that needs answering or a job that has to go out by five. Nobody is being difficult. They're being practical.
The new way needs to survive that pressure long enough to become the normal way. That's the real challenge, and most of it comes down to how you roll the change out rather than what the change is.
Pilot it before you roll it out
The biggest mistake I see is rolling a change out across the whole agency on day one. It feels decisive, but it means every team is learning something new at the same time, while still delivering the work.
Pick one team, one client or one type of job instead. Learn what works, fix what doesn't, then widen it. A small pilot is much easier to protect than an agency-wide launch, and the problems you find will be much smaller.
Give it an owner and a date
A change that belongs to everyone belongs to no one. Someone needs to be named as responsible for keeping it alive, and it helps if that isn't the busiest person in the building.
Set a date for the first checkpoint too, ideally within three or four weeks. Without a date, there's nothing to measure the change against, and nothing to make anyone ask how it's going.
Protect some time for it
If you expect people to learn a new way of working on top of a full schedule, it won't stick. It sounds obvious, but I see it all the time. The planner shows everyone booked at 100%, and the new process is expected to fit in the gaps. There aren't any gaps.
Block out a little capacity. Even an hour a week in the planner for the people involved sends a message that the change matters, and gives them space to get it wrong while they learn.
Run old and new side by side, then switch one off
For a while, you will have both ways of working in place. That's fine and often sensible. But it needs a limit.
Decide at the start how long the overlap will last, and set a date when the old way stops. If you don't, you end up with two systems running forever, double the admin and a team that never quite knows which one to trust. That is the point where most people give up on the new way and go back to the old one.
Show an early win
People need to see the change working within a few weeks. Not a big transformation, just something small and real. A handover that saved a round of amends. A weekly planning meeting that finished in 20 minutes instead of an hour. A timesheet report that showed up on time for the first time in months.
Share it. A single clear example does more to bring a sceptical team along than any amount of explaining why the change is a good idea.
Keep it visible
A change that nobody talks about is a change that's already fading. Give it five minutes in the weekly meeting. Ask the people using it what's working and what's getting in the way, then act on what they say.
It also helps to make sure the people at the top are using the new way themselves. If the founder or a director still asks for updates the old way, the team will quickly work out which one counts.
Know when to stop
This is the part people don't expect. Not every change deserves to survive.
If you've given it a fair go, with a pilot, an owner, some protected time and an honest look at the results, and it still isn't working, stop. Dropping a change that doesn't fit is better than letting it limp along for six months, using up goodwill and making the next idea harder to get started.
The aim isn't to force every change through. It's to give good ideas a proper chance and let go of the ones that don't earn it.
Why I don't stop at the roadmap
This is also how I work with agencies. I no longer offer an audit and roadmap on their own. Through experience, I know that recommendations rarely change anything by themselves. Without someone on hand to help put them into practice, they stay in the presentation deck while the agency gets back to the day job.
So I stay involved after the roadmap is agreed. That means helping to choose the pilot, agreeing who owns the change, checking in once the novelty has worn off and helping the team switch the old way off. It's the part most consultants leave out, and it's the part that decides whether the change sticks.
Changing the tyre without stopping the car
Agencies will always be busy, so there will never be a perfect moment to change how things are done. The ones that improve don't wait for the right moment. They start small, name an owner, protect a bit of time, set a date to switch the old way off, and keep the change in front of people until it becomes normal.
You don't need a perfect moment. You need a pilot, an owner and a date.
