How to Price a Freelance Retainer Without Losing Money
By RemoteSWE ·
Short answer: price the unit of work, not the month — and define the unit before you name a number. "Fifteen unique designs a month" and "three concepts resized five ways" look identical on a brief and are nowhere near each other in hours. Get that definition wrong and you have locked yourself into a year of underpriced work. When the client pushes back on price, remove units from the month rather than money from the unit, so your rate never moves. This post covers how to define the unit, the arithmetic on a real retainer, when a retainer is a trap, and how to review one before it quietly goes bad.
The short answer: define the unit first
| What the brief says | What it might mean | Hours apart |
|---|---|---|
| "15 unique designs" | 15 original concepts, or 3 concepts resized 5 ways | Roughly 5x |
| "Ongoing dev support" | Scheduled feature work, or unbounded on-call | Unknowable until priced |
| "A few revisions" | 2 rounds, or until they are happy | Infinite at the top end |
| "Manage the site" | Updates and backups, or updates plus content plus support | 2x to 4x |
Every one of these disputes is cheap to resolve before signing and expensive after. Ask the boring clarifying question. It is the highest-paid five minutes in the whole negotiation.
Why retainers go bad
A retainer is a bet that the average month looks like the month you imagined. The client is buying certainty; you are selling it. If your estimate of the unit is wrong by even 40%, you have sold certainty at a loss and agreed to keep selling it for the next twelve months.
The second failure is scope drift, which retainers are unusually vulnerable to because there is no project end to renegotiate at. Small additions accumulate month over month, each one too minor to argue about, until the retainer quietly contains twice the work it was priced for. That is the same mechanism as ordinary scope creep, but without the natural reset a fixed project provides.
The third is opportunity cost, and it is the one experienced freelancers actually complain about. Retainers absorb your capacity at a rate you agreed to months ago, while new clients pay today’s rate. A retainer that felt generous in January can be the reason you cannot take better work in June.
How to price one: the arithmetic
Work from the unit upward. Never from a monthly number downward, because a monthly number invites you to guess at what fits inside it.
- Define the unit precisely. One landing page. One original concept. One deployed feature under a stated size. Write it in a sentence the client could not reinterpret.
- Time the unit honestly. Use your slowest recent example, not your fastest. Retainers are priced on the bad month.
- Multiply by volume, then add the hidden hours. Calls, revisions, admin, context switching. For most retainers this is 20 to 30% on top and it is the number people forget.
- Apply your project rate, then a commitment adjustment. A modest discount for guaranteed volume is reasonable. Anything past 15 to 20% means you are paying the client for the privilege of being booked.
- Add a cap and an overage rate. The retainer covers N units. Unit N+1 is billed at the standard rate. Without this line the retainer has no ceiling.
Worked through: 8 landing pages a month, 4 hours each at your slowest, is 32 hours. Add 25% for calls and revisions and you are at 40. At $85 an hour that is $3,400. A 12% commitment discount lands the retainer at roughly $3,000 a month for 8 pages, with page 9 billed at $360. Every number in that chain is defensible, which is what you need when the client negotiates.
A worked example: how to handle "that is too expensive"
The client comes back and says $3,000 is above budget; they can do $2,400.
The move most freelancers make: keep 8 pages, drop the price to $2,400. The effective rate falls from $85 to about $68 an hour, permanently, for twelve months, and the client now anchors on that number for everything else you ever quote them.
The move that holds: keep the rate, take units out of the month. $2,400 buys 6 pages instead of 8. Same rate, smaller commitment, and the client still gets a monthly number that fits their budget.
This reframes the conversation from what you are worth to how much they want to buy, which is a far better argument to be having. It also leaves the door open: when they need the extra pages, they pay the overage rate rather than expecting them free. If they refuse to move on either volume or price, that is useful information about the account, not a negotiating setback.
When a retainer is a trap
There is a real counterargument to retainers and it deserves stating plainly: some experienced freelancers avoid them entirely, on the grounds that they consume the capacity you would otherwise sell to new clients at current rates.
- The scope cannot be bounded. "Available when we need you" is not a unit. Decline or convert it to a block of hours.
- It would take more than about half your capacity. One client at 50%+ is not a retainer, it is an employer with none of the protections.
- The rate is below what new clients currently pay. You are subsidising them with your own growth.
- There is no review date. An open-ended retainer is a rate freeze you agreed to by accident.
- The client wants it mainly for a discount. Retainers should buy predictability for both sides, not just cheaper hours for one.
How to review a retainer before it goes bad
Track three numbers per month and review them quarterly. This takes about ten minutes and prevents the most common way retainers turn into resentment.
- Actual hours versus priced hours. Drift above 15% for two consecutive months is a repricing conversation, not a personal failing.
- Effective hourly rate. Retainer fee divided by real hours, including calls. Compare it directly against what new clients paid you that month.
- Share of total capacity. Past half, start actively replacing rather than renewing.
Put a review date in the agreement from day one — every six months is reasonable. A scheduled review makes a rate increase a calendar event rather than a confrontation, which is the same reason new clients should always get the new rate first. The full mechanics are in how to set your rate and actually raise it.
FAQ
How much should I discount for a retainer?
Somewhere between nothing and about 15%, depending on how much certainty it genuinely buys you. Past 20% you are not being paid for predictability, you are paying for the privilege of being booked. If the client wants a larger discount, reduce the units instead.
Should a retainer be hours or deliverables?
Deliverables where the unit can be defined cleanly, hours where it cannot. Hours are safer for open-ended support work and worse for anything where you get faster over time, since efficiency then reduces your income rather than increasing your margin.
Do unused units roll over?
Standard practice is no, and you should say so explicitly before signing rather than after a light month. Rollover turns a capacity reservation into a bank balance the client can withdraw from all at once, which defeats the purpose.
Are retainers worth it at all?
They are worth it when they are bounded, reviewed and priced from the unit up. They are a slow-motion mistake when they are open-ended, discounted heavily and never revisited. The difference is entirely in the setup, which is why the definition conversation matters more than the number.
Define the unit, time your slowest version of it, add the hidden hours, then price from there — and when the budget does not stretch, take units out of the month instead of money off the rate. Put a review date in the agreement before you sign, and check your effective hourly rate against what new clients pay you every quarter. A retainer that is never reviewed is just a rate freeze with extra steps. For the pricing foundation underneath it, start with setting and raising your rate, and protect the scope with the change-request habit here. And if the positioning is what is holding the rate down in the first place, run it through the free profile scorer.