Agency ROI: deciding whether a retainer is actually worth taking

A retainer buys hours, so price it in hours

The headline on a retainer tells you almost nothing. It is a revenue figure, and revenue is the input to the calculation rather than the answer.

What a retainer actually sells is a slice of your team's capacity. Judging it means converting the money into a rate, and comparing that rate against what those hours cost you to supply.

• Effective hourly rate = (retainer − direct costs) ÷ hours actually delivered
• Direct costs are the ones that exist only because of this client: tools, ad spend, contractors, licences.
• Hours actually delivered includes the account management, the revisions and the reporting — not just the work you would put on an invoice.

The denominator is where retainers go wrong. It is almost always larger than the estimate, and the gap is almost always the unbilled work nobody recorded.

The hours nobody counts

Ask a team how long a client takes and you will get the production time. The real figure also includes:

• Account management and the internal stand-up where the client is discussed.
• Revisions, especially any round beyond the one you scoped, and especially when the brief was vague.
• Reporting and the monthly call where the report is presented.
• Ad hoc messages, and the context-switching cost of picking the work back up afterwards.

None of these feel like the work, so none of them get counted — and they are frequently a third of the total. A retainer that looks comfortable at production hours alone can be below cost once they are included.

The same retainer at two different hour counts

Illustrative figures. The point is the shape of the curve, not the specific values.

Nothing about the client changed between those two columns. The same money, the same invoice, the same team. A 58% increase in hours cut the effective rate by a third — and the extra 35 hours is exactly the unbilled work from the previous section.

If your cost of delivering an hour is above $31.58, the second column is a client you are paying to keep. That is a decision you are allowed to make, but make it deliberately rather than by accident.

You cannot sell all the hours you own

A person works roughly 160 hours in a month. Nobody sells 160 hours of client work, because the business also needs selling, admin, learning and slack between projects. A common planning target is 60–70% utilisation, which is about 96–112 billable hours per person per month.

This is why capacity, not demand, is usually the real constraint. When you accept a retainer that consumes 95 hours, you have not simply added revenue — you have removed a third of a person's month from every other opportunity.

The true cost of an underpriced retainer is not the money it fails to earn. It is the better-priced work you turned down because the calendar was full.

When to decline

Declining is not a growth strategy on its own. It is the necessary counterpart to pricing: a floor means nothing if you accept everything below it.

• The effective rate is below your floor rate, and no scope change moves it above.
• The scope is undefined. Undefined scope is not a discount problem, it is an uncapped liability.
• The client's process consumes hours you cannot bill — approval chains, changing stakeholders, meetings with no decision.
• The work is a loss leader with no named, dated follow-on. A loss leader without a path to the second sale is just a loss.

Four levers, in order of difficulty

1. Raise the price. Fastest to try, hardest to do, and usually available where the scope has quietly grown since the original agreement.
2. Cut the scope to what the price actually covers. Cheaper than raising the price, and it protects the effective rate the moment it is agreed.
3. Raise the floor. Apply the break-even rate to every new enquiry so the underpriced work stops entering the pipeline in the first place.
4. Productise the repeatable part. The retainer work that is identical for every client is the work to systematise, because it is the only lever that reduces the hours without reducing the output.

A quarterly review that takes ten minutes

An underpriced retainer does not announce itself. It arrives gradually, as extra calls that feel like service and then revisions that feel like scope creep, and by the time somebody adds up the hours the pattern is a year old.

Recalculating the effective rate once a quarter turns that into a decision rather than a discovery. Four numbers, ten minutes, per client.

• Hours delivered in the quarter, from a timesheet or a diary — not from memory.
• Direct costs attributable to that client.
• The retainer as invoiced.
• The effective rate, compared against the figure you agreed at the start.

Where the rate has fallen, the cause is usually visible in the hours: a new stakeholder, a weekly meeting that became two, or a deliverable that quietly expanded. Naming the cause is what makes the conversation possible, because a scope discussion needs a specific change rather than a general complaint about workload.

Work out your own break-even

The Agency ROI Calculator takes the retainer, the direct costs and the real delivered hours, and returns the effective hourly rate alongside the contribution after costs — so the decision is made against a number instead of a feeling about the client.