Signs Your Agency Needs a Project Manager

If you run a marketing agency and you are looking for signs your agency needs a project manager, the honest framing is narrower than the search suggests. The coordination work already exists in your agency. Someone is doing it, and in most agencies that someone is the owner, at the end of the day, badly. The only question worth answering is whether that work has grown into a job one person should hold, or whether it is still a set of habits you can fix without hiring anybody.

I have spent twelve years running delivery operations for marketing agencies across 92 completed engagements, so I sell the thing this article is about and you should read it with that in mind. What follows is deliberately limited to whether and when. If you already know you need the role, the separate question of how to find one and test them properly is covered in my guide to hiring a project manager, and nothing below repeats it.

Is the coordination work in your agency a role yet?

Three things have to be true at once before coordination becomes a seat rather than a habit. I would not hire against fewer than three, and in my experience the hires that disappoint fail on the third condition rather than the first two.

• Volume. There is enough coordination in a normal week to occupy a person for a meaningful part of it. Not a busy week. A normal one.
• Continuity. It is the same work every week, not a different emergency. A role can be defined around a repeating rhythm. It cannot be defined around whatever went wrong on Tuesday.
• Authority. You are prepared to let somebody else hold the schedule, including the unpopular calls: moving a date, pushing back on a client request, telling a strategist their draft is late. A project manager without the standing to do those things is an administrator with a task board, and you are still the project manager.

What can you check in ten minutes to find out?

None of this needs a consultant, and none of it depends on how the week felt. Most lists of signs are feelings written down, and feelings track how tired you were that week. Open the tools you already have and run these five checks in order, each of which produces evidence instead.

1. Count projects against producers. Write down how many client projects are genuinely live right now, then how many people are producing work on them. Coordination load lives in that ratio, not in your headcount. Six people running four projects have little to coordinate. Six people running twenty have a full-time job nobody has been given.
2. Count the handoffs in one typical project. Take a job you run often and count every point where the work changes hands: strategist to writer, writer to editor, editor to designer, designer to the person who publishes, and back to the client for approval. Coordination cost scales with handoffs. Two of them do not need a manager at any size. Eight of them do.
3. Read back your own sent messages from last week. Sort them into two piles: work only you could have done, and information you moved between two people who could have spoken directly. The second pile is a job description your own week has already written.
4. Try to name the promised date on your last five deliveries. Not the date they shipped. The date you committed to at the start, and where that commitment is written. If it is not recorded anywhere, you have found something more basic than a hiring question, and it is worth fixing before anybody new inherits it.
5. Recall the last week you were properly unavailable. Holiday, illness, a conference with no signal. If delivery carried on without you, coordination is distributed and you are earlier in this than you think. If everything waited for your return, coordination is concentrated in one person, and that person cannot also be the one selling and setting direction.

Read the five together rather than separately. One failed check is a bad week. Four failed checks that would fail again next month is a role.

Why is headcount the wrong trigger for this hire?

Most advice on this answers with a number of employees, and the number is doing very little work. Two agencies of the same size can have completely different coordination loads, because coordination is created by handoffs, dependencies and the number of separate clients waiting for something, not by how many chairs are occupied.

A twelve-person agency delivering one long retainer per client, with each account owned end to end by one person, may have almost nothing to coordinate. A six-person agency running short multi-stage projects for fifteen clients at once, where every project touches four people, is drowning in it. The second one needs the role first, and would be told to wait by every article that answers this question with a headcount.

This also explains why buying a better board rarely helps at this stage. A tool can display handoffs. It cannot chase one, decide which of two waiting people goes first, or tell a client the date moved.

When is the honest answer still not yet?

Four situations where the checks look bad and hiring is still the wrong move.

• You are measuring a peak. Three clients onboarding in the same month, or a launch quarter, produces every symptom of a structural problem and none of the permanence. A permanent hire made against a temporary peak becomes a fixed cost you carry through the quiet quarter that follows. Re-run the checks when nothing unusual is happening.
• Your delivery already runs in pods. If one person carries a client from brief to delivery, the coordination happens inside a single head and there is very little to hand over. Adding a project manager on top of that structure usually adds a reporting layer rather than removing work. What those teams need is a shared definition of done and one checkpoint a week, not a new seat.
• The real constraint is capacity, not coordination. If you have sold more work than your team can physically produce, a project manager will give you an accurate, beautifully maintained account of work that was never going to fit. The schedule will still be wrong. That is a pricing and hiring problem in production, and no amount of coordination rescues it.
• You have never tried the cheapest version. If your agency has never run a single weekly delivery checkpoint with a written agenda, you do not yet know whether you need a role or a rhythm. Some of what feels like a missing person is a missing thirty minutes.

What to do instead for the next four weeks

If any of those four apply, here is the interim version. It costs nothing, and it either fixes the problem or proves the role is real.

1. Run one thirty-minute delivery checkpoint a week, same time, written agenda: every live project, one owner, one next action, one risk. No status theatre, no round-the-room updates.
2. Record the promised date for every project at the point you commit to it, in one place, in a form somebody other than you can read.
3. Name one contact per client so that information stops travelling through you by default.

Run that for four weeks and take the five checks again. If they still fail with a rhythm in place, you have your answer, and the four weeks cost you nothing you would not have spent anyway.

What does waiting too long actually cost?

Waiting has a cost too, and it is quieter than the cost of hiring early. The bill does not arrive as a crisis. It arrives as decisions you stop making.

• You stop selling. The pipeline goes quiet during delivery-heavy months because the person who sells is the person holding the schedule together.
• You start declining work you could deliver. Not because the team is full, but because you cannot face the coordination the new account would create.
• Improvement stops. Nothing gets reviewed after it ships, so the same avoidable problem is solved from scratch every quarter by whoever is nearest to it.

So does your agency need a project manager?

My verdict, with the bias declared once more: hire when volume, continuity and authority are all true, when a typical project crosses more than a handful of hands, and when the checks fail in an ordinary month rather than a brutal one. Do not hire on headcount, and do not hire because the last quarter hurt. If the checks fail and you are ready, the remaining question is the shape of the seat rather than the decision itself: a full-time employee, a fractional project manager, or the coordination function bought as a service. That choice, and how to test a candidate properly, is covered in the guide to hiring one. My basis for all of this is twelve years running delivery operations for marketing agencies and 92 completed engagements, which is the kind of record I would ask anyone to produce before you take their advice on when to hire.

Where to start if you are not sure

The five checks are deliberately something you can run alone, and for most owners that is enough to reach a decision. When it is not, the Agency Ops Audit exists for exactly that gap: $1,500, two weeks, a written 90-day roadmap and a 60-minute readout call, built to answer whether coordination is your binding constraint at all and what it would cost you to keep absorbing it. If the answer is that you are not ready for this hire, that is a finding, and I would rather deliver it than sell you a seat you cannot fill.