Agency operations

Retainer

A recurring fee for an agreed scope of ongoing work.

A monthly arrangement covering either a defined set of deliverables or a block of hours. It gives the agency predictable revenue and the client predictable access, and it is the model most marketing agencies are built on.

Key takeaways

  • Deliverables caps the agency exposure and invites arguments about what counts. Hours caps the client exposure and makes every inefficiency visible.
  • Value-based pricing works where the agency genuinely controls the outcome, which in marketing is rarer than proposals suggest.
  • Retainers fail at the edges, not in the middle. The work nobody priced is the work that erodes the margin.
  • Pricing month one setup into the ongoing fee makes every later month look overpriced.
  • Review the scope on a schedule, including when nobody has complained. That is the review that prevents the crisis one.

Three structures, and what each one protects

The choice of structure is a choice about who carries the risk of the work taking longer than expected, and it is worth making on that basis rather than on which is easiest to sell.

  • DeliverablesA fixed set of outputs per month. Caps the agency exposure to a known list, and invites an argument about what counts as one of them. Best where the work is repeatable and the client is reasonable.
  • HoursA block of time, drawn down as used. Caps the client exposure instead, and makes every hour of inefficiency visible and chargeable to somebody patience. Best where the work is genuinely unpredictable.
  • Value-basedA fee tied to the outcome. The hardest to sell and the hardest to hold, because the client product, pricing and sales team all sit between the work and the result.

Retainers fail at the edges

The core of a retainer is usually fine. Whatever was named in the agreement gets delivered, the team knows how long it takes, and the margin on it is roughly what was modelled.

What erodes the account is everything adjacent: the quick question that becomes a half-day, the extra asset for a channel nobody priced, the report format the new marketing manager prefers, the meeting that was going to be fifteen minutes. None of it is in the agreement and none of it is refusable in the moment.

That is why the total hours on an account can drift a long way from the priced hours without anybody making a decision. There was never a moment where somebody agreed to do more work; there were forty moments where somebody agreed to do a small thing.

The structural response is to make the edges visible rather than to police them. Hours tracked against the account, reviewed monthly, converts an invisible drift into a number, and a number can be discussed. Without it the first hard evidence arrives as a margin review or a resignation.

The first month, and the shape of the fee

Onboarding is front-loaded and the fee usually is not. Audits, access, tracking setup, strategy and the first round of assets happen in month one, and then the ongoing work settles into something considerably lighter.

Folding that setup effort into a flat monthly fee produces a specific and predictable problem: month one is underpriced, and every month from four onward looks expensive relative to what the client sees happening. The client is not wrong about that, which is what makes the conversation hard.

Separating them fixes it. A setup fee for the onboarding work and a lower ongoing retainer describes what is actually happening, and it makes the ongoing number defensible for as long as the relationship lasts.

It also protects against the early exit. An agency that priced six months of setup recovery into the monthly fee and loses the client in month three has funded the onboarding out of its own margin.

Reviewing the scope before anybody is unhappy

Scope is written once and reality moves continuously. A retainer agreed in January describes a client who had different priorities, a different team and a different budget from the one you have in September.

The habit that works is a scheduled review, quarterly, held whether or not anything is wrong. Three questions: what is in the scope that nobody uses any more, what is being done that was never in it, and what should the next quarter be about.

Held on a schedule, that is a normal business conversation and often a commercial opportunity, because the honest answer usually includes work the client needs and is not buying. Held only when something has gone wrong, the same conversation is a negotiation under pressure with a client who has already decided something is unfair.

The unused hours question belongs here too. Rolling unused capacity forward indefinitely sounds generous and creates a liability: a client who has banked three months of hours can call on all of them in a week, at the worst possible moment, entirely within their rights.

What it means for an agency

Write down what is included, what is extra, and how often the scope is revisited, then revisit it on that schedule even when nobody has complained. The review you hold while things are fine is the one that prevents the conversation you would otherwise have in a crisis.

Where it goes wrong

  • Rolling unused hours forward indefinitely, until a client has banked three months of capacity they can call on at once.
  • Pricing the first month setup effort into an ongoing monthly fee, which makes every later month look overpriced.
  • Not tracking hours on a deliverables retainer, which leaves the drift invisible until it is large.
  • Selling value-based pricing on outcomes the client product and sales team actually control.

In short

A retainer is a bet that the work will average out, and it goes wrong through accumulation rather than through any single decision. The agencies that hold their margin are not the ones that say no more often; they are the ones that can see the drift while it is still small.

Frequently asked questions

Should a retainer be based on hours or deliverables?

It depends on who should carry the risk of the work taking longer than expected. Deliverables caps the agency exposure to a known list and invites arguments about what counts as one. Hours caps the client exposure and makes every inefficiency visible. Predictable, repeatable work suits deliverables; genuinely unpredictable work suits hours.

How do you stop a retainer becoming unprofitable?

Track hours against the account even when the agreement is deliverables-based, and review the total monthly. Retainers do not fail in the middle, they fail at the edges, where forty small unpriced favours add up without anybody deciding to do more work. Hours make that drift into a number, and a number can be discussed before it becomes a margin review.

Should onboarding be included in the monthly fee?

Better to price it separately. Setup work is front-loaded, so folding it into a flat fee underprices month one and makes every month from four onward look expensive relative to what the client sees happening. A separate setup fee describes what is actually happening and keeps the ongoing number defensible.

Should unused retainer hours roll over?

If they do, cap it. Indefinite rollover sounds generous and creates a liability: a client who has banked three months of capacity can call on all of it in one week, at the worst possible moment, and they are entitled to. A one-month cap, written down at the start, is the usual compromise.

Related terms

Last reviewed

Keplent keeps track of these for you.

It holds what each client agreed to, watches what their campaigns actually did, and tells you in plain sentences what needs you this week.

Join the waitlist