One person's capacity is measured in hours, not in logos. You work it out by dividing the billable hours actually left in their month, once holidays, internal meetings and everything that never becomes client work have been taken out, by the hours an average client costs with its coordination counted inside.
Why counting logos always gets it wrong
Almost every small agency carries capacity in somebody's head, expressed in accounts. One person has five, another has seven, and when a new one comes in it goes to whoever looks least underwater. That works until two of those accounts grow at the same time, and then it is not a client that fails: it is the person, and they fail on every client at once.
The trouble is that a logo is not a unit. An account with twelve carousels a month and one contact who approves the same day, and another with eight videos and four people commenting by email, are written identically on an org chart and occupy completely different months. Counting logos is counting the wrapper.
The correct unit is the hour, and there are two things to count separately. How many hours an average client costs per month, including everything that is not production. And how many billable hours a person genuinely has in a month, which is not their working schedule. Dividing one by the other gives capacity, and it usually lands well below what anybody would say from memory.
One person's real month
This is where most estimates break. Somebody starts from a full working day, multiplies it by the days in the month and gets a round number that exists nowhere. The month has to be reduced, and each subtraction has a reason.
| Taken out of the month | Why | What happens if you skip it |
|---|---|---|
| Holidays and sick days | Spread the year across twelve months | August looks like the only odd month and it is not |
| Internal meetings | They are paid for and not billed | Half a day a week that appears on no sheet |
| Proposals you do not win | They are part of the work | They get charged against the wrong client |
| Training and tooling | They happen inside the working day | They become hours outside the working day |
| The gap between one account and the next | Switching context costs | It is the difference between the sheet and the month |
The first two subtractions almost everybody accepts. The last three are the ones that stop an honest calculation from looking like a plan, because they are what turns a full working day into something noticeably shorter than anyone wants to write into a proposal.
It is worth spreading holidays and sick days across all twelve months rather than treating them as a summer hole. Bad days do not concentrate in August, they concentrate wherever they land, and a model that piles them into one month returns eleven optimistic months and one disaster.
What a client costs, hour by hour
The second half of the division is the one people fill in with the visible part of the work. Production, and not much else. But production is, on small accounts, less than half of what a client actually consumes.
| What a client costs | Where it hides |
|---|---|
| Production | Pieces per month times hours per piece, caption already written |
| Rework | The share of what was produced that gets touched again after a change of mind |
| Community | Replying, moderating and listening, every day and with no fixed slot |
| Reporting | Collecting the metrics, building the report and explaining it |
| Calls | The meeting plus the time afterwards writing down what was said |
| Chasing approvals | Reminding, resending and asking whether they have seen it |
| Admin | Brief, access credentials, invoices and files that never arrive |
Of those seven lines, the one most people set to zero and almost nobody actually has at zero is rework. It is a share of production rather than a loose hour: if a third of the pieces get touched again, production does not cost what the sheet says, it costs a third more on that part.
And two lines appear in no proposal in this industry while happening every single month: chasing approvals, and the small admin around a client. Added to the calls and the report, on a small account they take a slice of the month comparable to producing. That figure is what explains how somebody with five quiet clients can be more loaded than somebody with eight well organised ones. It is also the one line item a client never sees on an invoice and never believes when it is described to them, which is why it is worth measuring it for a month rather than arguing about it.
Do the division with your own numbers
The sum only means anything with your hours rather than somebody else's, so this guide is not going to hand you a clients per person figure. What there is instead is a clients per person calculator that does both halves: you describe an average client across its seven lines, describe a real month with its subtractions, and it returns what the client costs, what billable hours are left and how many fit.
It is worth running twice. Once with the most loaded person on the team, which is where the risk is. Once with the newest, because their month looks nothing like it: less autonomy, more questions and more revisions, and treating them like the first person is how teams lose new hires in the second quarter.
One warning about what that page counts and what it does not. It counts hours and people, never money. If what you want to know is what to charge for that month, that is a different sum, and it is done by the social media pricing guide and the monthly rate calculator.
Everything in this article happens in one place in GoFeed.
Try it freeThe whole number, and the fraction left over
The division almost never lands on a round number, and that is the interesting part. If it comes out with decimals, plan against the whole number below it. Rounding up is the decision that burns a person out, because the fraction left over is not a small client: it is the month's margin.
It is worth looking at from the other side. That fraction is what absorbs the week when two calendar deadlines collide, the day a platform changes something, the three day absence and the client who decides at the last minute that the campaign starts earlier. All of that happens, and it happens every month on one account or another. If the fraction has already been sold, it gets paid for with hours that do not exist, and it is always the same person who pays.
The second use of that figure is more useful than the first. It is not only how many fit, it is how many they carry today and how much room is left, and that comparison is what turns capacity into a hiring decision instead of a complaint. A capacity that is overshot two months running is not a streak: it is a team sized wrongly that has not owned up yet.
What the sum does not see
No division settles this on its own, and it is worth saying where it falls short before making a hiring decision with it on the table.
- Clients are not interchangeable. A video account and a carousel account can land on the same hours figure and not fit in the same month, because video concentrates and carousels spread out.
- The onboarding curve. A new client costs far more in its first two months, and the division treats it as though it were already at cruising speed. Adding two new accounts in the same month is where most small teams come apart.
- Concentration. Capacity can be comfortable across the month and short in the week when several calendars close at once. The month adds up and the person does not sleep.
- The cost of switching context. Nobody performs the same split across many accounts as focused on a few, and that loss fits in no box. It is a real argument for concentrating accounts per person even when the split looks less even.
What you can recover before hiring
Before adding a person, look at the lines that are not production, because they are the only ones you can bring down without bringing down the quality of what goes out. Producing faster has a floor; chasing things less does not.
The two most recoverable hours in that table are chasing approvals and rebuilding the monthly report, and both come back the same way: the client has somewhere to look and approve without being reminded, and the report prepares itself. In an agency workspace that means a portal per client, an approval where one side saying yes is enough, and a monthly report that goes out on the last working day. How many clients fit afterwards is still your division, and you will not find a multiplier here saying otherwise.
The third recoverable line is the small admin, and that one is fixed with order more than with tools: access credentials requested once and properly, files somewhere the client can reach on their own, and the month's tasks written where the whole team sees them instead of in one person's head. How that gets organised is covered in tasks, chat and Drive, and the shared workspace as a whole on the team module.