Capacity Planning for Agencies: How Much Work Can You Really Take On?

A client asks whether you can start a new project in three weeks. You look at the calendar, you look at your team, you feel a vague sense of dread, and you say yes anyway. Two weeks later Koos is working evenings and a project that was supposed to be quiet has eaten the whole month. I have done this. Most agency owners I know have done this. The problem is not that we are bad at maths. The problem is that the answer to "can we take this on" lives in four places at once and none of them talk to each other.

Capacity planning is working out how much work your team can realistically absorb in a given period, so you can decide what to accept, what to push back and when to hire. For agencies and freelancers with several clients, it is the difference between a good month and a month you survive. And almost every guide you will find on the subject assumes something you probably do not have: a clean record of how everyone spent their hours.

How do you do capacity planning without timesheets?

You can do useful capacity planning without tracking a single hour, as long as you plan in committed days per person instead of billable hours per project.

Here is the version that works for a small team. For each person, write down how many working days they genuinely have available in the coming weeks, after holidays, admin, sales calls and the standing client meetings that never move. That number is almost never five per week. For most people in a small agency it lands between three and four, and if you plan on five you are overbooked before the first project starts.

Then put the work next to it. Not tasks, not tickets, just the chunks of work you have committed to, with a rough size in days and the name of the person who has to do it. If a project needs Bep specifically, it does not matter that Sjef has room. Availability that sits with the wrong person is not capacity.

You will see a benchmark quoted a lot in this corner of the internet: aim for roughly 70 to 80 percent billable utilisation. It is a reasonable target for a firm with a resource manager and a timesheet culture. For a five person agency it is close to useless as a planning instrument, because you cannot measure it accurately enough to steer on it, and chasing it tends to push you into accepting work you should have refused. Plan in days you have promised.

Why does capacity planning break at agencies?

Capacity plans fail at agencies because the biggest variable, the work that has not been sold yet, sits in a different system from the work that has.

This is the part the template articles skip entirely. Your planning board knows about the six projects you won. Your pipeline knows about the four proposals out with clients, one of which is a near certainty and starts in two weeks. A capacity plan built only from signed work tells you that you have room. You do not have room. You have room until Thursday.

The practical consequence is that a capacity plan is only as good as the connection between sales and delivery. If those live in two tools, someone has to remember to carry the information across, and under pressure nobody does. I wrote more about that gap in why a normal CRM is not enough for an agency. The fix is not a better spreadsheet. It is making sure the proposal you sent on Monday shows up in the same view as the work you are already doing, with a probability and a likely start date attached.

Three other things that quietly wreck a capacity plan:

• Planning per client instead of per person. Ten boards for ten clients cannot answer "is Truus overloaded next week". That is a different exercise, and I covered how to structure it in keeping an overview of multiple clients.
• Treating the plan as a document. A capacity plan you rebuild once a quarter in a spreadsheet is a snapshot of a week that has already passed.
• Ignoring the maintenance work. Retainers, small change requests and the client who calls every Friday are real capacity. If they are not in the plan, the plan is fiction.

What should you actually track?

Four things, and you can hold all of them in a simple tool: available days per person, committed work per person, expected work from the pipeline with a start date, and the point in the future where those three cross.

That last one is the only output that matters. Not a utilisation percentage, not a colour coded heat map. A date. "We are full until 12 October, and if the Aurelius proposal lands we are full until early November." That sentence is what lets you tell a client honestly when you can start, quote a realistic deadline, and see a hiring decision coming two months out instead of two weeks out.

Where software helps, and where it does not

Dedicated capacity planning tools exist and some are good. Resource Guru, Float and the resource modules inside Monday or ClickUp all give you a grid of people against weeks. If you have twenty people and a full time resource manager, buy one.

For a team of two to ten, my honest view is that a separate capacity tool becomes another system nobody updates. What you need is for your existing project view to also know about your pipeline, which is the reason we built planning and the CRM into the same product in Wecooking rather than shipping a standalone capacity module. Being straight about the limits: Wecooking has no time tracking and no invoicing, so if your capacity planning depends on comparing planned days against logged hours, we are not the tool for that, and this piece on time tracking next to a CRM explains when you genuinely need it.

The uncomfortable truth is that most agencies do not have a tooling problem here. They have a habit problem. Fifteen minutes on a Monday, going through people rather than projects, beats any software you buy and do not maintain. Start there. If after a month you want the pipeline and the plan in one screen, that is when a tool earns its place.

You can try Wecooking for free and see whether having sales and planning in one view changes your answer to "can we take this on". From 19 euros per month, everything included, and currently free in beta.