What you are actually trading
A retainer is a recurring monthly fee in exchange for ongoing work or availability. The usual framing is that you accept a lower rate in return for certainty, which is true but incomplete.
The full picture includes something freelancers routinely leave out of the comparison: the unpaid hours that project work requires. Finding clients, writing proposals, having exploratory calls that go nowhere, negotiating, and onboarding each new client are all real hours that nobody pays for.
A freelancer billing $100 an hour on project work who spends eight hours a week on business development is effectively earning considerably less per working hour than the headline figure suggests. A retainer at $85 an hour with almost no sales overhead can genuinely pay better per hour worked, even though it looks like a discount.
That is the real trade: a lower nominal rate for a much higher proportion of billable hours. The guide to billable hours covers how to measure your own figure, which is what decides whether the trade is good for you.
The two structures, and why one is riskier
Almost every retainer is one of two things, and they behave very differently.
The access retainer
The client pays for availability: a set number of hours or days per month, or simply the right to call on you. Common in consulting, advisory and support work.
The risk is that availability has no natural ceiling. If the agreement does not define hours, "available" quietly becomes "responsive to anything, immediately", and the effective rate collapses. Access retainers need an hours cap and a stated response time, not because clients abuse them deliberately but because nothing in the arrangement creates a limit.
The deliverable retainer
The client pays for a defined output each month: four articles, eight social assets, one report, a monthly maintenance window. Common in content, design, SEO and development maintenance.
This is the safer structure for the freelancer because the boundary is built in. The work is countable, so overrun is visible immediately rather than at the end of a bad quarter.
Where you have the choice, prefer a deliverable retainer, or an access retainer with a hard hours cap, which amounts to the same thing.
What discount is reasonable
Ten to twenty percent against your standard project rate is the normal range. The justification is real: guaranteed income, no sales cost for that revenue, no repeated onboarding, and a client who already understands how you work.
Below ten percent, clients often do not perceive enough benefit to commit. Above twenty, you are usually giving away more than the certainty is worth, particularly if the retainer occupies time you could fill at full rate.
Two things should reduce the discount rather than increase it. If the retainer requires you to hold specific days open, that is a real constraint on your other work and deserves compensation. And if the work is unpredictable in volume, you are carrying risk the client would otherwise carry, which is worth a premium rather than a discount.
The clause that causes most retainer disputes
Not the price. Whether unused hours roll over.
The client's natural assumption is that hours they paid for and did not use should carry forward. Your natural assumption is that you reserved the capacity and the month is gone. Both positions are reasonable, which is exactly why it needs deciding in advance.
The common resolutions:
- No rollover. Cleanest for you, and defensible on the grounds that the client is buying reserved capacity. State it plainly at the outset.
- Limited rollover. Unused hours carry to the following month only, then expire. A fair middle ground that most clients accept readily.
- Rollover with a cap. Hours bank up to a ceiling, commonly one month's worth. Reasonable for lumpy work where volume genuinely varies.
Whichever you choose, put it in writing before the first invoice. This single sentence prevents more retainer friction than everything else combined.
When project work is the better business
Retainers are not automatically superior, and the cases where project work wins are specific.
- When your rate is rising quickly. A retainer locks in today's rate. Early in a career where you are repricing every six months, that lock is expensive.
- When the work is genuinely finite. Some projects end. Stretching them into a retainer creates make-work that neither side values.
- When one client would dominate your income. A retainer worth more than about a third of your revenue is a concentration risk that behaves like employment without any of the protections.
- When the value is front-loaded. A rebrand, a strategy, an architecture. These are worth more as projects than amortised across twelve months.
The mix most established freelancers end up with
The arrangement that tends to work is neither all retainer nor all project. A common shape is retainers covering roughly half of baseline costs, with project work above that.
The retainers mean your fixed costs are met before the month begins, which removes the pressure that causes panic-priced work in quiet periods. The project work keeps your rate current, brings in new clients, and prevents any single relationship becoming structural. It also means a retainer ending is a setback rather than a crisis.
If you are building toward that, the budgeting guide covers how to work out what your baseline actually is.
Review it every six months
Retainers drift. The scope expands gradually, the monthly fee does not, and after eighteen months you are delivering half again as much for the same money. This happens to almost everyone and almost nobody notices in the moment, because each addition was small.
Schedule a review every six months, and say at the outset that you do. Track actual hours against the agreement, and if the gap has widened, bring the number up. A scheduled review is a routine conversation; an unscheduled one after two years of drift is an awkward one.