How many hours should a freelance retainer include?

At most the monthly fee divided by your floor rate. A 3,400 retainer against a floor of 85 covers 40 hours, and that number is a ceiling rather than a goal. What makes a retainer different from any other fixed fee is not the arithmetic, which is the same division as always. It is that the arrangement renews. A fixed-price project can only go underwater once and then it is delivered. A retainer goes underwater and comes back on the first.

A retainer sells availability, so the hours have no edge

A project has an edge built into it. There is a deliverable, it ships, and the hours stop whether or not anyone was counting them.

A retainer has a date instead. What the client buys is that you are there this month: the quick question on Tuesday, the call that was going to be fifteen minutes, the small change that was easier to do than to scope. None of those look like work worth logging on their own. Together they are the month.

This is why retainer hours drift without anyone deciding to change anything. There is no moment where a deliverable is renegotiated, because there is no deliverable. The agreement on paper is exactly what you signed. The month underneath it is not the month you signed up for.

Your retainer has an hour ceiling

The ceiling is the same division as a fixed-price hour budget, applied to a month instead of a project.

Take the monthly fee. Divide by your floor rate, the lowest rate at which you still want the work.

3,400 ÷ 85 = 40 hours.

Forty is not what you should aim for. It is where the client stops clearing the line you set for everyone else. At 41 hours this retainer pays worse than any other client you would accept, and it does so quietly, because the invoice is identical every month.

Notice what the ceiling does not depend on: the number of hours in the agreement. Plenty of retainers say “up to 30 hours” and then run at 45. The written number tells you what was promised. The ceiling tells you what the fee can afford. When those two disagree, the fee wins, because the fee is the part that actually arrives.

Three months, one fee, three rates

Here is what the drift looks like when you do the division every month instead of once.

Month	Hours	Fee	Effective rate
Month 1	34	3,400	100
Month 2	40	3,400	85
Month 3	50	3,400	68

Nobody sent a change request. Nobody had the conversation. Month two is the interesting one: 40 hours is precisely the ceiling, so the client is paying exactly the least you were ever willing to accept. That is the month to notice, and it is the month that looks completely normal from the outside.

Why this is worse than a bad fixed-price project

A fixed-price project that runs over is a bounded loss. You finish it, you learn the real hours, and the next quote starts from the real number. The damage has an end date.

A retainer has no end date by design. That is what the client is buying and it is usually what you want too, because predictable income is the reason to take one. The same property is what makes a bad retainer expensive: month three at 68 an hour is followed by month four, and the reason nothing changes is that nothing looks like it changed. The invoice is the same. The client is happy. The only place the drift exists is in hours nobody added up.

That is also why a retainer can sit at the top of your revenue list and the bottom of your rate list at the same time. Ranked by fee it is your best client. Ranked by the hour it can be your worst, and the two orders rarely match.

What to do before the renewal date

Divide the fee by your floor and write the number down. One number, once. 3,400 over 85 is 40. Put it where you will see it on the tenth, not in a spreadsheet you open in January.

Read three months, not one. A single heavy month is noise: a launch, a bad week, one client emergency. Three in a row is not noise, it is the current scope, and the scope is what you are actually selling.

Count the hours that do not feel like work. Calls, Slack, the small request that took twenty minutes and one context switch. Those are the hours a retainer runs on, and they are the ones missing from a timesheet reconstructed on Friday. They are also non-billable in name only here, because the retainer is supposed to cover them.

Change it at the renewal, not mid-month. Bring the ceiling and three months of hours. “The last three months ran 34, 40 and 50 hours against a fee that covers 40” is arithmetic. The same sentence in week three of a bad month is a complaint, and it is much easier to say no to.

Decide which lever moves. More fee, fewer hours, or a narrower scope. All three are legitimate. Absorbing it is also a choice, as long as it is one you made on purpose and not one you discovered in March.

Where CronLoom comes in

CronLoom has no retainer mode, and it does not send invoices. What it has is the division, kept current.

You set the month’s retainer up as a project with a fixed fee, and the effective rate for that project becomes fee divided by every hour logged against it rather than hours times a rate. Set the project’s budget hours to the ceiling you calculated, and the 80% alert lands at hour 32 of 40. That is mid-month, while the month can still go differently, which is the whole point. At 100% you hear about it again.

The effective hourly rate calculator does the same division for a single month if you want to check last month before you change anything. If you want it tracked as the hours happen, join the waiting list.

Do it for last month

Open last month. Add up every hour that went to your retainer client, including the calls and the quick questions. Divide the fee by that number.

Then look at the date the retainer renews, and decide whether the number you just calculated is one you would sign again.