Agency overhead ratio
Overhead ratio is your non-delivery cost as a share of gross income: the money left after freelancers, media and pass-through spend. It sets the floor under every rate you quote. An agency that grows overhead faster than gross income gets busier and poorer at the same time.
Written for founder-led UK agencies of roughly £500k–£2M. Overhead rarely arrives in one decision. It accumulates through reasonable choices made a quarter apart.
What counts as overhead
Overhead is every cost that is not delivering client work: leadership and new business salaries, finance and administration, premises, software used across the agency, and the non-billable share of anyone who also delivers. Sort costs once, apply the rule consistently, and the trend becomes readable.
| Cost | Where it belongs | Why it gets misfiled |
|---|---|---|
| A director who bills two days a week | Split: billable share to delivery, the rest to overhead. | Filed entirely as delivery, which flatters margin and hides the cost of management. |
| New business and marketing your own agency | Overhead. | Treated as an investment outside the ratio, so it never gets reviewed against what it wins. |
| Project management software, design tools | Overhead unless bought for one client. | Accumulates per seat as headcount grows, and nobody revisits the licence count after someone leaves. |
| Freelancers and media spend | Neither. Strip out before calculating gross income. | Left in revenue, which makes the ratio look healthy on a pass-through heavy month. |
Where it accumulates
Hires made ahead of the revenue
An agency wins a large account and staffs for the version of itself it expects to be. The account ends or shrinks; the salary stays. This is the single largest cause.
Seniority added instead of structure
Decisions keep reaching the founder, so the agency hires a senior person to absorb them. Without written decision limits, the decisions keep reaching the founder anyway, and the payroll carries the cost.
Software bought per problem
Each tool solves something real and costs little on its own. Counted together across two or three years, the total surprises most founders.
Space taken for a headcount you no longer have
Premises get chosen against a growth plan. When the plan changes, the lease does not.
Holding it flat without cutting delivery
The fastest correction is to freeze overhead for two quarters and let gross income catch up, rather than cutting into the capability that earns it. Cutting delivery to fix an overhead problem removes the capacity you need to grow out of the problem.
Plot eight quarters
Overhead and gross income on one chart. The point where the lines diverge names the decision that caused it, and you will usually remember making it.
Audit the licences
Seats against people, subscriptions against use. The quickest saving available to an agency, and the least painful.
Give every hire a revenue trigger
Agree in advance the contracted income that justifies the next hire. Write it down, and the argument happens once rather than every quarter.
Fix the decision load before adding seniority
Install written decision limits first. Some agencies then find they do not need the hire, and the ones that do get more from it.
Common questions
What overhead ratio should an agency run at?
Work backwards from your rate card instead of borrowing a figure. At your day rates and realistic utilisation, calculate the gross income your billable team can produce, then the overhead that leaves the margin you intend to keep. That number is your ceiling. Published averages come from agencies with different rates, salaries and service mixes.
Should the founder's salary sit in overhead?
Split it the same way as any other part-billable person: the share of time spent delivering client work goes to delivery, the rest to overhead. Leaving a founder's cost out entirely is the most common reason an agency believes it is more profitable than it is, and it distorts every rate calculation downstream.
We need to cut. Where should we start?
Software licences and subscriptions first, since they cut without losing capability. Then premises at the next break point. Touch delivery capacity last, because rebuilding it costs more than the saving and takes longer than the quarter you are trying to fix.
Is a high overhead ratio always bad?
No. An agency investing in new business or building a product carries deliberate overhead and should judge it against what it returns. The problem is overhead nobody chose, accumulated through hires and tools added a quarter at a time. Sort the ratio into chosen and inherited, and the review becomes straightforward.
Related reading: Fractional COO for agencies · Agency utilisation rate · Agency project and retainer margin · Busier than ever, and the bank balance disagrees
See where the lines diverged
Bring two years of overhead and gross income. The quarter they separated usually points straight at the decision behind it.
Book an Operational Clarity Call →