How do I decide whether a retainer makes sense?

How do I decide whether a retainer makes sense?

I look for ongoing work that only I can do, a decision the client needs made regularly, and a scope I can describe in one sentence. If the work is really a queue of unrelated tasks, a retainer is just a discount on chaos and I would rather quote projects.

Retainers get recommended as the answer to freelance income instability, and they can be. They can also be the thing that quietly eats your calendar while producing the least interesting work in your practice. Which one you get depends almost entirely on how the arrangement is defined before it starts.

I price on a fixed fee and most of my projects land between one thousand and ten thousand dollars, so a retainer has to compete with well defined project work rather than with empty weeks. That comparison keeps me honest.

What is a retainer actually buying, for each side?

For the client it buys availability and continuity, which means not re-explaining their business every time something needs doing. For me it buys predictable capacity planning. Neither side is buying a fixed number of hours, and pretending otherwise is where most retainers go wrong.

The moment a retainer becomes an hour bank, the relationship changes. The client starts counting, I start defending, and both of us are managing a spreadsheet instead of the outcome. Hours are an input, and nobody hired me because of my inputs.

The version that works is defined by responsibility. I own this outcome, in this area, continuously. That framing survives a busy month and a quiet one, and it keeps the conversation on whether the outcome is being delivered.

When does a retainer genuinely fit?

When something needs a decision every month and getting it wrong is expensive. Ongoing AEO and GEO work fits, because search and answer engines change and content needs judgement continuously. Running automations in production fits, because they need monitoring and someone to own the failures.

The automations I keep in production are the clearest example. An Airtable and WhaleSync pipeline or a HubSpot flow through Zapier is not finished when it ships. It needs someone watching for the day an upstream system changes shape, and that responsibility cannot be bought as a one off project because the risk is continuous.

Content and visibility work fits for the same structural reason. There is a decision every cycle about what to publish, what to update, and what to cut, and the value comes from the accumulation rather than from any single deliverable. Continuous work also needs a visible heartbeat, which is why I send a short written summary at the end of every week rather than waiting for a monthly report.

When do I say no to one?

When the client wants a cheaper way to get unlimited requests, when nobody on their side owns the relationship, or when the work is genuinely finished. A site that is built and stable does not need me every month, and saying so is better business than inventing maintenance.

The unlimited requests version is the most common and the most polite. It arrives as flexibility. Can we just have you available for whatever comes up. That sounds reasonable and it produces a scope that grows silently until one of us is unhappy, which is the same dynamic as ordinary scope creep on projects with a monthly invoice attached.

The ownership version is subtler. If nobody on the client side is responsible for making decisions about the work, a retainer becomes a slow queue where my questions wait for weeks. I have learned to ask who I will be talking to every week before agreeing to anything ongoing, because the answer predicts whether the arrangement will work.

And if the honest answer is that the work is done, I say that. Telling a client they do not need a retainer has never cost me a relationship. It has produced referrals more than once.

How do I price it without guessing?

I price the responsibility, not the hours, by estimating a normal month and a bad month and charging closer to the bad one. If a bad month would make me resent the arrangement, the price is wrong, and no amount of goodwill fixes a price that was set for the best case.

A practical way to get there is to write down what happens in a normal month, then write down the worst realistic month. A platform change, an automation breaking, a launch moving. The gap between those two is the risk I am carrying, and the fee has to cover carrying it.

I also price so that the client can leave. Long lock-ins protect the wrong thing. If the work is worth paying for each month, monthly terms are fine, and if it is not, a twelve month contract just delays an honest conversation. This is the same reasoning that led me to compare hourly billing against fixed scope sprints and land on defined scope in both models.

What goes in the agreement so nobody resents it?

Four things. The outcome I own, what is explicitly outside it, how requests arrive and get prioritised, and what response time means in practice. Everything else is detail, but a retainer without those four turns into a negotiation every month.

The exclusions matter more than the inclusions. Writing down that a new page template or a new integration is separate project work is not a way to squeeze money out of a client, it is how both of us keep a stable understanding of what the fee covers. Clients who are told this up front rarely mind. Clients who discover it in month three reasonably do.

Prioritisation is the one people skip. A single channel where requests arrive, and a rule about what gets done first, prevents the situation where the loudest request beats the most valuable one. I would rather agree that rule when nobody is stressed.

What does a retainer need to survive month four?

Visible value and a changing agenda. Month one is easy because there is a backlog. Month four is where retainers die, usually because the obvious work is done and nobody has decided what the next quarter is for.

My fix is to review the point of the arrangement every quarter rather than every year. What changed, what is next, and whether the fee still matches the responsibility. That conversation sometimes ends a retainer, which is a feature. The alternative is an arrangement neither side believes in that continues out of politeness.

Reporting helps only if it reports on the outcome. A monthly summary of tasks completed teaches a client to value tasks. A monthly summary of what moved, what I decided, and what I recommend next teaches them to value judgement, which is what they are actually buying.

How do I end one well?

By making the handover real. Documentation of what exists, access transferred, a short recorded walkthrough of anything unusual, and a clear statement of what will need attention without me. An ending done well is the best marketing I have.

This matters especially for automations, because an automation nobody understands becomes a liability the moment its author leaves. If I built it, someone on the client side should be able to read what it does, know where it logs, and know how to turn it off.

I also try to leave the door open without being needy about it. Clients come back when their situation changes, and the ones who came back had good exits. That has been true across more than seventy projects for over twenty five clients in six years of this, and it is the pattern I would protect above short term revenue.

What is the honest downside of retainers?

They make your calendar predictable and your work narrower. Ongoing responsibility crowds out the experiments and the harder one off projects, and if every slot you have is committed, the interesting inbound enquiry arrives at the worst possible time.

I feel this tension constantly. Continuous work is comfortable and it compounds knowledge of a client's business, which produces better decisions. It also means saying no to work that would teach me something new, and that trade is easy to get wrong in the direction of comfort.

So I keep a deliberate limit on how much of my capacity is committed, and I revisit it. Capping the roster rather than maximising it is the decision behind why I capped my retainer roster, and I still think a smaller number of ongoing relationships done properly beats a longer list done thinly.

What should you do next?

If a client has asked you for a retainer, write down the outcome you would own, the worst realistic month, and the three things explicitly outside it. If you cannot fill in all three, you are not ready to quote, and the conversation you need is a scoping one.

If you are on the other side of this and wondering whether to put someone on retainer, apply the same test. Is there a decision every month that you need made well. Reach out if you want to talk through whether your situation actually calls for one.