Is a marketing agency worth it for a 20 person firm?

A managing partner at a 20 person accountancy practice in the UK has a proposal on the desk. £1,800 a month on a twelve month term, for search and paid advertising. The partner group meets and the conversation goes where it always goes: can we afford £1,800 a month, and what did the last agency ever do for us.

Neither question can be answered. The first is the wrong figure, because a twelve month term commits the practice to £21,600 and the monthly number is only the payment schedule. The second is a memory rather than a test.

The answerable question is narrower. £21,600 has to come back as profit, not as revenue, and it has to come back from clients the practice does not have yet. So how many of those clients, and for how long, does £21,600 take? That is arithmetic, and it needs two figures the practice already holds plus one it almost certainly does not.

The comparison most firms run is against the wrong number

£21,600 of fee is not covered by £21,600 of new fee income, because most of that revenue leaves again as the cost of delivering the work. A practice that wins £21,600 of new billing and keeps 55 per cent of it after fee earner time, software, support and partner review has kept £11,880. Rather more than half the agency fee is still outstanding, and the engagement looks like it broke even.

The figure that decides it is contribution: what a client leaves behind once the cost of serving them has gone out. Comparisons published on this subject almost always set the fee against turnover instead, which flatters the fee by the whole of the firm's delivery cost. It is the same defect that makes a quoted range useless: a number with no scope behind it is not a price, and a return with no margin behind it is not a return.

Break-even on one client, expressed in years

Divide the annualised fee by the annual contribution of a single client, and the answer arrives as a length of time rather than as a verdict. The figures below are illustrative. Replace all three with your own and the method holds.

Corporation tax relief does not move the break-even

The fee is deductible, and so is nothing else in the calculation, because the income the agency produces is taxed at the same rate. At the 25 per cent main rate, £21,600 of fee costs £16,200 after relief. The £2,310 of contribution is taxed on the same terms and becomes £1,732.50. Divide one by the other and the answer is 9.35 years, which is the figure the untaxed sum already gave. HMRC sets the main rate at 25 per cent on profits above £250,000 and the small profits rate at 19 per cent below £50,000. Marginal relief covers the band between. Whichever band a practice sits in, both sides of the sum move together. A supplier who nets the fee down for tax without netting the return down has shown you half a sum.

The input most firms cannot state

Average client life decides the answer, and it is the one figure a partner group rarely has to hand. The method is one division on a year of your own records: clients at the year end, divided by clients lost during that year. A practice with 400 clients that lost 40 has an average relationship of ten years, and a break-even of 9.4 clears comfortably on a single client.

Where a sector publishes the raw counts, the same division works on them. The FCA's 2025 retail intermediary data records 4,132,298 ongoing client relationships at the year end and 286,874 that ceased during the year at retail investment adviser firms. That divides to an implied average relationship of about fourteen years. The regulator does not publish that number, the division produces it, and advice relationships are unusually long, so an accountancy or recruitment practice should expect a shorter figure. The transferable part is the method, not the fourteen. It also explains why the cost of a lead tells an advice firm so little on its own: the value it is compared against arrives over more than a decade.

How the answer moves on your own two figures

Below roughly £2,000 of average annual fee, one client cannot carry a retainer of this size at any plausible margin. Each cell is the number of years a single new client has to stay to cover £21,600 of fee. Divide by the number of clients you expect.

When a 20 person firm should not hire an agency yet

Two situations make the sum unanswerable, and neither is about the agency quoting.

• Your fee earners are already at target utilisation. New work then needs a hire before it needs marketing, and a salary carries employer National Insurance and pension contributions on top of it. None of that sits in the retainer arithmetic, and a practice that wins nine clients it cannot staff has bought a problem rather than a return
• You cannot state your average client life. Break-even is a period, so without a retention figure there is nothing to compare it against, and the decision reverts to the argument you were trying to escape. The division takes an hour and it settles the question for every supplier you ever quote

We are paid by the agency, so a page that talks a partner group out of a retainer costs us money. What we assess before recommending anybody, including the cases where we decline to introduce anyone, is set out on the front of this site.