A marketing retainer is the most common way to buy marketing in South Africa and one of the least understood things a business pays for. You agree a number, the number leaves your account on the first, and what comes back is a handful of posts, maybe a report, and a general sense that things are happening. Ask what the money bought and you will usually get a list of activities rather than an account.
That is not because agencies are dishonest. It is because a marketing retainer is structurally a block of hours sold as a monthly fee, and almost nobody explains the hours. This is what is inside them.
A retainer is hours, not deliverables
The first thing worth knowing is that you are not buying four blog posts and eight social posts. You are buying a share of a working month, and the deliverables are whatever that share happens to produce. The distinction sounds academic until the month a website migration eats the time that would have gone on content, and you get two posts instead of four with no explanation, because nothing in the paperwork ever said four.
It is also why the same rand figure buys very different things at different firms. A R15,000 retainer at an agency with a R450 effective hourly rate is about 33 hours. At a firm carrying more overhead it might be 18. Neither will print that number, and the difference between them is most of what you are actually choosing between.
| What you are told | What it usually means |
|---|---|
| A dedicated account manager | Someone who attends the meetings, and is not the person doing the work |
| Ongoing optimisation | Whatever hours are left after the committed deliverables |
| Monthly reporting | Two to four hours, and often the most automatable line on the invoice |
| Strategy included | A plan written once at onboarding and referred to rarely afterwards |
| Unlimited revisions | Revisions come out of the same pool, so they cost you output somewhere else |
Where the hours actually go
Across our own accounts a month tends to split into four parts. Production is the part clients picture, and it is rarely more than half of it.
- Production, meaning the things you can look at: 40 to 55 per cent.
- Research and planning, meaning keyword work, competitor checks and deciding what to make and why: 15 to 25 per cent.
- Measurement and reporting, meaning pulling numbers, checking they are real, and writing them up: 10 to 15 per cent.
- Coordination, meaning approvals, revisions, chasing assets, meetings and admin: 15 to 25 per cent.
The uncomfortable part of that list is that about half the month produces nothing you can look at. That is not waste. Research is what stops the production half being spent on the wrong thing, and measurement is the only reason anyone can say afterwards whether it worked. But it does explain why a breakdown is rarely volunteered: half the answer sounds like overhead until it is explained, and explaining it opens a conversation most agencies would rather not have.
About half of a retainer produces nothing a client can look at. That is not waste. It is just never explained, so it looks like it.
The coordination tax, and who pays it
Coordination is the line that varies most between agencies, and the one most worth interrogating, because you are partly in control of it. Every approval round, every asset that arrives late, every meeting held to reach a decision that could have been an email, comes out of the same pool as the work.
A client who approves in two days rather than two weeks genuinely gets more marketing for the same money. Not as a reward, but because the hours that would have gone on chasing go somewhere else instead. Few agencies say this out loud because it sounds like blaming the client. It is worth saying anyway, because it is the only lever on this list that costs nothing to pull.
What most people get wrong
The common mistake is treating the monthly fee as the thing to negotiate. An owner pushes an R18,000 proposal down to R15,000 and feels they have won, when what has happened is that the hour allowance dropped and nobody changed the deliverables list. The scope no longer fits the hours, so the gap gets absorbed by whichever part of the month is least visible. That is almost always research and measurement, which is to say the two things that would have told you it was going wrong.
The second mistake is judging the work by volume. Four posts a month is not better than two if the two were aimed at something and the four were made to fill a calendar. Volume is the easiest thing to measure and the least connected to any outcome, which is exactly why so much of it ends up written into contracts.
Two questions that change the relationship
You do not need to become an expert in agency economics. Two questions do most of the work, and both are reasonable to ask before signing anything.
- 01How many hours a month is this, and what is the scope in hours rather than in deliverables? A firm that cannot answer does not track it, which means nobody there knows whether you are being over-serviced or under-serviced. A firm that can answer and will not is telling you something too.
- 02What happens when we go over? The honest answers are that it gets flagged before the work is done, or that it is absorbed up to a point and then discussed. The answer to be wary of is a vague one, because in practice it means the work quietly stops and nobody tells you which part.
The point
None of this asks an agency to work differently. It asks it to say what it is already doing. A scope in hours, a statement of hours used, and an honest answer about what happens at the ceiling would settle most of the arguments this industry has with its clients. None of it is difficult. It is only uncomfortable, and comfort is the reason it does not happen.
Sources and method
- South African agency retainer bands cross-checked against published rate cards and recruiter salary surveys, August 2026. No figure here comes from our own client work.
- Hour-split ranges are this workspace own logged time across five accounts since April 2026, stated as a range and labelled as ours rather than as an industry benchmark.
Written by Assay House, a South African brand strategy and creative agency. Published 28 August 2026.
