Agency Capacity Planning: How Much Work Your Team Can Take Before You Hire

Agency capacity planning is the practice of working out how many sellable hours your team actually has in a period, how many of those hours are already committed to existing retainers and projects, and what the remaining gap is worth in revenue at your real hourly rate. For a marketing, creative or digital agency, it answers one question: can we sign this retainer without breaking delivery? This guide produces two numbers you can calculate this week — sellable hours available next quarter, and the revenue those hours can carry — and then uses them to choose between four outcomes: hire, subcontract, reprice, or say no.

This is not shift scheduling. It is not about who covers Saturday or how many people are on the rota at 6pm. It is about whether you can sign the client, and what happens to your margin if you do.

By Ed Kamel, founder of Zerak and Wieldy. Wieldy began as the internal system of the marketing agency he runs, so he writes from running retainers, sales commissions and media buying himself. Published September 2026. Competitor prices re-checked quarterly; next review December 2026, reviewed by the Wieldy team.

Why is agency capacity not the same as headcount?

Because headcount tells you how many people you pay, not how many hours you can sell. A 12-person agency running at 40% utilisation and a 12-person agency running at 78% are completely different businesses with the same payroll. The first has roughly double the room to take on new work; the second is one sick week away from missing deadlines.

Start by splitting the roster into three groups:

Four things then quietly eat what is left:

• Internal meetings. Standups, all-hands, one-to-ones, client calls that were never scoped.
• Pitch and proposal work. Every unpaid pitch is billable-grade time spent on a client you do not have yet.
• Rework and revision rounds. The third round of amends on a social set is capacity, whether you charged for it or not.
• Account management on "delivery-only" retainers. Status calls, reporting, chasing approvals. Scoped as zero, delivered as hours.

There is one trap specific to agencies. Retainers create a standing monthly claim on your hours before a single project is booked. If you calculate capacity project-first, you will keep discovering that the retainers already spent it. Calculate retainer-first, project-second, always.

Step 1: How do I calculate my sellable hours?

Take contracted hours for the period, subtract time off, then subtract a fixed non-billable allowance per role. What remains is what you can actually sell.

The baseline hours formula

For each role group:

Sellable hours = (working days in period × hours per day) − (holiday + sick + public holidays) × target utilisation for that role

Run it per role group, not for the agency as a whole. Averaging a media buyer and an office manager into one number produces a figure that is wrong for both.

Setting a realistic utilisation target by role

Utilisation rate is the share of available hours that is billable. The ranges below are the ones I use in practice as an agency owner, and I am labelling them as practitioner judgement rather than a published industry benchmark — I am not going to quote you a figure I have not measured.

A delivery person at a sustained 95% is not efficient. They are a person with no slack, and slack is what absorbs the client emergency you cannot predict.

Worked example: a 10-person agency

All numbers below are illustrative — they are not Wieldy data and not a benchmark. Team: 6 delivery staff, 2 account managers, 1 owner, 1 admin. Period: one 13-week quarter.

Base availability per person, for illustration:

• 13 weeks × 5 days = 65 working days
• Minus 5 days holiday, 2 days sick allowance, 2 public holidays = 56 working days
• 56 days × 7.5 productive hours = 420 available hours per person per quarter

Now apply role targets:

Ten people, 4,200 hours of paid time, 2,352 sellable hours a quarter — about 56% of the total. That gap between 4,200 and 2,352 is the single most common error in agency planning. Owners take contracted hours at 100%, conclude that everything fits, and find out in week three that it does not.

Step 2: How much of that capacity is already committed?

List every active retainer and every project in flight, convert each to hours per month, and subtract. Use delivered hours where you have them, not contracted hours — in most agencies the two differ, and the difference is your real capacity.

Retainers: the hours you have already sold

For each retainer, write down the hours the contract implies and the hours you actually delivered last month. If a retainer is scoped at 20 hours and consistently consumes 31, your capacity plan must use 31. Anything else is fiction.

That gap has a name: retainer scope creep, and the overservicing it causes. Most agencies treat it as a billing problem — "we should have charged for that." It is a capacity problem first. An overserviced retainer silently removes hours you had already promised to someone else, and you will not see it until a deadline slips on a different account.

Rebuild this table monthly:

(Illustrative figures.)

Projects in flight and projects sold but not started

Phase project hours across the weeks they will actually be worked. A $30,000 build sold in January does not consume its hours in January. Dumping the full estimate into the month of sale makes that month look impossible and the following months look empty — the exact opposite of the truth.

For an agency, the month a project lands matters more than its total size. Two comfortable projects that both need their design phase in the same fortnight will break you; the same two projects staggered by three weeks will not. Use the budget at completion calculator to sanity-check whether a project in flight is tracking to its estimated hours or quietly overrunning them.

Subtract committed hours from sellable hours. What is left is your gap. In the illustrative example above, 2,352 sellable hours per quarter is roughly 784 a month; committed hours of 31 + 38 + 60 = 129 leaves a large nominal gap — which is why the next step matters, because hours alone will not tell you whether to take the work.

Step 3: What is the gap worth in money?

Multiply spare hours by your effective hourly rate, not your rate card. Your effective hourly rate is revenue delivered ÷ hours delivered over the last completed quarter. It is almost always lower than the rate on your proposals, because the rate card does not include the amends, the status calls or the hours you wrote off.

Illustrative: $180,000 delivered last quarter ÷ 2,100 hours delivered = $85.71 effective hourly rate. If your next-quarter gap is 400 hours, your revenue capacity is roughly 400 × $85.71 = $34,284. That is the ceiling on what you can sell without changing something.

Now run it in reverse for the retainer on the table. A prospective $6,000/month retainer ÷ $85.71 = 70 hours a month, or 210 hours a quarter. Does your 400-hour gap cover it? Yes — with room for one more mid-sized client, and nothing else.

Then check contribution per hour: the revenue the work brings in per hour, minus the direct delivery cost of that hour. A retainer that fills your capacity at below your effective rate makes the agency busier and poorer. You lose the ability to say yes to better work, and you gain no margin for it. That is the worst trade in agency management, and it is made every week.

Two tools to run the actual numbers on the actual deal:

• Agency profit calculator — model the margin on the specific retainer before you sign it.
• Retainer pricing calculator — find the price at which the work becomes worth the capacity it consumes. Background reading: how to price agency retainers.

Both are free, with no sign-up, and each shows its formula with a worked example.

Step 4: Should you hire, subcontract, reprice or decline?

Run the breakeven first. If the committed pipeline cannot cover a new hire's fully loaded cost for several months, the answer is not a hire — even if the next month looks busy.

The breakeven a new hire has to clear

The fully loaded cost of an employee is salary plus employer taxes and benefits, software licences, equipment, recruitment, and the management time the role consumes. Wieldy does not publish a recommended loading percentage for this, because employer costs on top of salary vary by country and by market — work out your own multiplier from your actual payroll costs rather than borrowing one from somewhere else.

Illustrative, using a $60,000 salary and an illustrative 25% loading for employer costs and benefits (set your own figure based on your market):

Expected billable hours: 1,680 available hours a year (illustrative), at a 75% target = 1,260 billable hours.

$82,000 ÷ 1,260 = $65.08 per hour, just to break even. At a 30% target margin on that role, the hire needs to generate roughly $93 an hour — comfortably above the $85.71 effective rate in our example. Which tells you something uncomfortable and useful: at your current rates, that hire loses money unless your rates go up or their utilisation exceeds 75%.

Three months of runway, not one

Do not hire against one month of overflow. One busy month is noise. Before hiring, I want to see committed revenue — signed retainers and signed projects, not proposals out — covering the fully loaded monthly cost of the role for at least three months, plus a ramp allowance.

Ramp is the part everyone forgets. A new delivery hire rarely reaches target utilisation on day one; assume six to twelve weeks of reduced billable output while they learn your clients, tools and standards. That ramp is a real cost and it lands before any of the revenue does.

When a freelancer or contractor is the right answer

Use a freelancer when the demand is spiky, when you need a skill twice a quarter, or when the client driving the overflow has not yet proven they will renew. A freelancer converts fixed overhead into variable cost, and that is exactly what you want when you are not sure the work persists.

The cost is real, though, and it is not just the rate. You give up margin per hour, you take on quality control and briefing time, and your institutional knowledge walks out at the end of the engagement. For a three-month overflow on one account, that is a good trade. For work that will still be there in a year, it is not.

When the right answer is a price rise or a no

If you are sustainably above your utilisation ceiling and turning work away, the market has told you your rates are low. The fix is not another body — it is a price rise on new work, then on renewals. Model it in the retainer pricing calculator before you send the email.

And sometimes the answer is no. Saying yes at 105% capacity is how delivery quality collapses, and the client you overserviced to win is rarely the client who forgives you for missing the deadline. Three ways to decline without losing the relationship:

• Defer the start. "We can take this on properly from the first week of March. We can't do it justice before then."
• Phase the scope. Take the highest-value component now and schedule the rest.
• Refer it on. A good referral to a trusted peer buys more goodwill than a mediocre delivery.

How often should the capacity plan be updated?

Monthly, in about 30 minutes, on a rolling 90-day horizon. An annual capacity plan is out of date by February.

The monthly review has four steps:

1. Update delivered hours per client for the month just closed.
2. Re-forecast the next 90 days — retainers first, then phased project hours.
3. Flag every client over or under service against contracted hours.
4. Check the hiring signals and decide: hire, freelance, reprice, decline, or do nothing.

Three early-warning signals are worth acting on:

• Utilisation above your agreed ceiling for two consecutive months. One month is a spike. Two is a pattern.
• Rising rework. Revision rounds climbing usually means the team is rushing, which is a capacity symptom before it is a quality one.
• Delivery dates slipping without scope changing. The clearest evidence that the plan and reality have parted company.

Keep the horizon at 90 days. For an agency whose pipeline is measured in weeks, anything beyond a rolling 90-day forecast is a guess with a spreadsheet around it. Pair this with the numbers in your agency profit margin tracking so the capacity review sits alongside the rest of your monthly numbers.

When does dedicated resource planning software beat a spreadsheet?

When allocation conflicts start costing more than the licence — which for most agencies is somewhere above roughly 15 delivery staff. Below that, a well-maintained sheet plus a disciplined monthly review usually beats a tool nobody updates.

Dedicated resource planning and scheduling tools go genuinely deeper than any spreadsheet. Productive.io and Scoro both include schedulers inside broader agency platforms, and Teamwork.com adds capacity planning from its Accelerate tier. If you are scheduling dozens of people across many concurrent projects with skill-based matching, a dedicated tool can be worth it. The spreadsheet will lose at that scale.

The hidden cost is that all of them charge per user, so adding a scheduler on top of everything else scales with your headcount:

All prices as listed on each vendor's own pricing page, checked September 2026 — prices may change.

I am not going to tell you any of these tools lacks a feature. Feature sets move quarterly; check with the vendor for anything that will decide your purchase. What I will say is that at ten delivery people, a spreadsheet rebuilt monthly using the steps above gives you the same decision quality for nothing.

Where does Wieldy fit in capacity planning?

Wieldy is not a minute-level resource scheduler, and I would not recommend it as one. Wieldy is agency management software where the money side of capacity lives: retainers and their contracted value, projects and tasks for what is committed, invoicing, payroll and HR records, and the profit picture that tells you whether a hire pays for itself.

In practice, that means:

• Retainer records showing contracted value and scope per client — the starting point of any committed-hours table.
• Projects and tasks for what is already promised and to whom.
• Payroll and HR records for the fully loaded cost side of the breakeven calculation.
• The built-in AI assistant, which answers questions about your agency's own live data by text or voice — who owes money, profit this month — in English or Arabic, including Egyptian and Gulf Arabic.

Pricing is flat per workspace, not per user, with no per-client fees. That matters for capacity planning specifically, because adding the freelancer or the new hire does not increase your software bill mid-seat-tier.

Launch pricing is 40% off regular ($99 / $175 / $290) and is locked for as long as the subscription stays active. The offer ends 18 October 2026. Yearly billing is 10× the monthly price: $590, $1,050, $1,740.

There is a 7-day free trial on every plan with no credit card required (14 days when signing up through a partner link), and a live demo workspace with sample agency data that opens from wieldyapp.com with just an email — no account. If you have existing data in spreadsheets or another tool, talk to the team about moving existing data before you start.

All the calculators referenced in this guide are at free agency calculators — free, no sign-up, and embeddable on your own site.

Frequently asked questions

How do I work out how much work my agency can take on?

Calculate sellable hours per role group (available hours × a realistic utilisation target), subtract the hours already committed to retainers and phased project work, then multiply the remaining gap by your effective hourly rate. That gives you both a capacity in hours and a ceiling in revenue. Compare the prospective client's monthly fee divided by your effective rate against the gap.

What is a good utilisation rate for a marketing agency?

In my experience running an agency, delivery specialists sit around 70–80%, senior delivery and creative directors 50–65%, account managers 30–50%, and owner-operators 20–40%. These are practitioner ranges, not a published industry benchmark. A sustained figure above 90% for delivery staff is a warning sign, not an achievement — it means no slack for the emergency you cannot forecast.

How much revenue does a new hire need to generate to break even?

Divide the fully loaded cost by expected billable hours at target utilisation. For illustration only: a $60,000 salary plus an illustrative 25% employer-cost loading (set your own figure — it varies by country), $3,000 of software and equipment and $4,000 of recruitment gives $82,000; at 1,260 billable hours a year that is about $65 an hour just to break even, and closer to $93 at a 30% target margin. Compare that against your actual effective hourly rate.

Should I hire a freelancer or a full-time employee for overflow work?

Use a freelancer when demand is spiky, the skill is needed occasionally, or the client causing the overflow has not proven they will renew — it converts fixed overhead into variable cost. Hire full-time when committed revenue covers the fully loaded cost for at least three months and the work will still exist in a year. The freelancer's real cost is margin and quality control, not the day rate.

What is the difference between capacity planning and resource scheduling?

Capacity planning asks whether the agency can take the work at all and what it is worth — hours, revenue and margin over the next 90 days. Resource scheduling assigns named people to named tasks on named days. Capacity planning is a commercial decision made monthly by the owner; scheduling is an operational one made weekly by delivery leads. Dedicated per-user scheduling tools specialise in the second.

Do I need resource planning software for a 10-person agency?

Usually not. Below roughly 15 delivery staff, a spreadsheet rebuilt in a 30-minute monthly review gives the same decision quality as a per-user tool, at no cost. Buy a dedicated scheduler when allocation conflicts across people and projects start costing you more than the licence — and remember that Productive.io, Scoro, Teamwork.com, Function Point and Workamajig all price per user, so the cost scales with headcount.