Every agency owner has run this calculation on the back of an envelope, and almost every envelope is wrong in the same direction: it counts the salary and ignores everything wrapped around it.
There are only three ways to add delivery capacity. Each one fails differently, and the failure is predictable enough to choose around.
Option 1: hire in-house
The obvious move, and the right one eventually. The problem is timing. A mid-level specialist costs the salary plus 25–40% in employment overhead, plus eight to twelve weeks before they bill at full rate, plus the senior time spent getting them there.
- Breaks when: demand is lumpy. You pay for peak capacity in every trough month.
- Hidden cost: the person who trains them is your most billable person.
- Best for: the one discipline that defines your positioning and that you will never outsource.
Option 2: freelancers
Fast, flexible, cheap on the invoice. The invoice is not where freelance capacity costs money.
- Breaks when: three of them have to coordinate, or your best one takes a full-time job in the middle of a launch.
- Hidden cost: management load. Someone on your team becomes a project manager for people who do not share your process, your tools or your reporting format.
- Best for: genuine spikes and genuinely specialist one-offs.
Option 3: a capacity retainer
A delivery team that already has the process, the reporting and the specialists, working under your brand against a fixed block of monthly capacity. You buy specialist days rather than job titles, and decide during the month what they get spent on.
- Breaks when: the partner has no NDA, no single channel and no fixed reporting format. Then you have bought a larger version of the freelance problem.
- Hidden cost: the discipline required to write proper briefs. The model rewards agencies that plan a month ahead and punishes the ones that do not.
- Best for: disciplines you sell constantly but do not want to define you — and any month where you are choosing between turning work away and hiring in a panic.
Turning down a signed project is not caution. It is a capacity decision you already made months earlier by not building the option to say yes.
Why the third option changes the arithmetic
Hiring and freelancing both leave the utilisation risk with you: you pay for the idle days, either as salary or as a scramble to keep someone booked. A capacity retainer moves that risk to the supplier, who spreads the same people across several partners and therefore runs them at 85–90% rather than 70%.
That difference is the entire economics of the model. It is also why a retainer can cost less per unit of finished work than a freelancer whose day rate looks cheaper on paper.
Hire what defines you. Freelance what surprises you. Put on retainer what you sell constantly, deliver repeatedly, and would rather not rebuild a team around every eighteen months.
A cleaner way to decide
Stop comparing day rates. Ask four questions instead:
- Is this discipline part of why clients choose us, or part of what they expect us to have?
- Is demand for it steady enough to fill twelve months, or does it arrive in bursts?
- If the person doing it left tomorrow, how many client relationships are exposed?
- How much senior time will be consumed managing whichever option we choose?
Most agencies end up running all three at once. The ones with healthy margins are simply deliberate about which work goes where — and they decide before the contract is signed, not after the client asks when the work starts.
