Project work pays like weather: some months it rains, some months nothing, and the forecast is a shrug. A retainer converts one good client relationship into something closer to climate — a known monthly number you can put in a cash-flow forecast without wincing. For a one-person business, one or two retainers covering baseline expenses changes the texture of every other decision you make.
But a retainer sold at the wrong time, or structured vaguely, is worse than project work. It's a subscription to guilt. Here's when it makes sense, the three structures that actually work at solo scale, and the terms that keep one from quietly consuming your life.
When a retainer makes sense
Two conditions, both required. First, a genuinely recurring need: the client has work that arrives every month whether or not anyone plans it — maintenance, content updates, reporting, small improvements, the operational drip that never finishes. A one-off build wearing a retainer costume is just a payment plan with extra anxiety. Second, trust already earned. A retainer is a bet on the relationship, and clients make that bet after you've delivered, not before. Don't propose one on a first engagement. Finish a project, be good to work with, and let the retainer become the obvious next shape of a thing you're already doing.
The three structures
1. Hours bank. The client prepays a block of hours each month — say ten hours at your rate — and draws it down. Simple to explain, easy to track, and the natural first retainer. Its weakness is that it prices you as a commodity: the client watches the meter, and every task gets mentally compared against the hours remaining. Fine for maintenance work; less fine when the work is judgment-heavy.
2. Capacity reservation. The client isn't buying hours; they're buying your availability — a guaranteed slice of your week, first claim on your calendar, a named response time. Price it like an insurance premium, because that's what it is: they pay so that when something breaks, you are theirs. This structure pays best and demands the most discipline about scope, because "availability" expands to fill any boundary you don't write down.
3. Care plan. The productized version: a fixed monthly fee for a defined list of recurring services — updates, backups, monitoring, a monthly report, up to two small tasks. Because the list is fixed, the plan is easier to sell and easier to refuse extras against: anything not on the list is a project. For a solo operator this is often the sweet spot — it reads to the client like a product, not like renting a person.
Pricing: discount for predictability, never for exposure
A retainer justifies a modest discount — commonly 10–15% off your project rate — and the reason matters: you are selling certainty. The client pays reliably on the first of the month; in exchange they get a better rate and priority scheduling. You can state that trade in a forecast and a proposal in the same sentence. What never justifies a discount is vagueness: exposure, future volume, "lots of work coming," a logo for your portfolio. Those are not currency, and they price at full rate forever. If a prospective client leads with future volume, the correct rate is your current one.
Anchor the math in your own numbers before you quote anything. What do you need per month to stop worrying, and what fraction of your week are you actually selling? A retainer that covers 30% of your target revenue for 20% of your hours is excellent. One that covers 40% of revenue for 70% of your hours is a job you gave yourself, with worse benefits.
Rollover policy
Two defensible options. Use-it-or-lose-it — unused hours expire at month's end — is the cleaner one, and it isn't a gotcha: the client is reserving your capacity whether or not they use it, the same way a gym membership doesn't refund January because you were busy. The alternative, one-month rollover, lets unused hours carry into the following month and no further. It feels friendlier and costs you little, because the one-month cap prevents the accumulation that kills retainers — the client who banks forty hours and then spends your entire December. What you never offer is unlimited rollover, which converts your retainer into an interest-free loan of your future time, callable at the worst possible moment.
The SLA you can actually keep as one person
Response-time promises are where solo retainers go to die, because they're written on a good day and tested on a bad one. Write the SLA for your worst plausible week, not your average one: next-business-day response for standard requests, same-day acknowledgment for genuinely urgent ones, with "urgent" defined as the site being down or money visibly on fire. Do not promise nights or weekends. A client who needs 24/7 coverage needs a team, and saying so is positioning, not apology. Then do the one thing that makes any SLA credible: keep it. A retainer where you answer in four hours every time beats one where you promised one hour and deliver a guilty six.
Exit terms
Thirty days' written notice, either side, no penalty, no exit interview, no guilt clause. This is the term that makes everything else work. The client stays because the arrangement is useful, not because leaving is expensive — which is precisely why they sign in the first place. For you, it means a client gone cold or a workload gone sour ends with a month of paid runway instead of a slow-motion standoff. Some operators ask for sixty days from the client side to protect the forecast; fair if you can get it, but keep your own exit at thirty. Asymmetry in your favor is the one negotiating perk of being small.
How to propose one
The best moment is right after a project goes well, when the recurring need is visible to both of you. Keep it to three sentences:
Now that [project] is live, you'll have the usual monthly drip — updates, small fixes, the reporting we talked about. I hold a retainer spot for exactly this: [structure and price], which covers [list] and gets you priority scheduling over project work. If it's useful I'll send the one-page agreement; if the drip turns out smaller than it looks, no harm — we stay project-by-project.
No deck, no discovery call about the discovery call. You're offering to formalize something already happening, at a better rate, with an easy no attached. About half of the right clients say yes; the ones who don't were telling you the need wasn't actually recurring, which is information you wanted anyway.
If you want the paperwork ready-made, the Client-Ready Business Templates ($19) include the retainer agreement with the rollover, SLA, and exit language above, alongside the SOW and change order for the project side. For the operational half — which recurring tasks are worth scripting so a retainer doesn't eat your week — see the automations piece. The structure is the point, though: pick one of the three, write the boundaries down, and let the first of the month become the least interesting day of your calendar.