Hourly contracts and the work diary

Hourly is for work where the scope will move: ongoing development, maintenance, anything you want to direct as it goes rather than specify up front. It runs on a weekly cycle, and understanding that cycle removes most of the confusion.

The weekly cycle

1. A buffer is funded, covering the maximum the coming week can cost.
2. The freelancer works and logs hours in the work diary as they go.
3. The week closes.
4. You review what was logged and approve.
5. Approved hours are released from the buffer to the freelancer.
6. The buffer is topped back up from your card for the following week.

The cycle repeats until one of you pauses or ends the contract.

The weekly hour limit

When you agree an hourly contract you set a rate and a weekly hour limit. The limit is the control that stops open-ended work becoming an open-ended bill.

The buffer funded each week is the rate multiplied by that limit. That is the most a week can cost you, and it is committed before the week begins.

Set the limit to what the work genuinely needs. Too low and the freelancer hits the ceiling mid-week and stops; too high and you are committing more than necessary each week, though you only ever pay for hours actually approved.

The limit can be raised later if the work grows, which funds the difference at that point.

The work diary

The freelancer logs hours against the contract with a note describing what the time went on. You see it as the week progresses, not only at the end — worth checking mid-week rather than discovering a surprise at approval.

The notes are the substance. Hours logged against "development" tell you nothing; hours logged against specific work tell you what you are paying for. If notes are thin, ask early. Most freelancers will improve them immediately; it is not usually deliberate.

Approving the week

At the end of the week you approve. Weeks are aligned to a fixed boundary rather than to when the contract started, so every contract closes on the same day and everyone knows where they are.

Approve everything if the work and the hours make sense.

Approve fewer hours than were logged if something does not. This is a real option and it exists for a reason.

But use it carefully. Cutting hours without saying anything is how working relationships end. If a week looks wrong, ask first — there is usually an explanation, and if there is not, the freelancer would rather hear it than be quietly paid less. Somebody who logs in good faith and is silently cut will not be around for the next contract, and will say so in their review.

If you do nothing, the week auto-approves after a period, in the same way milestones do. This protects a freelancer who worked from a client who stopped responding. If you are unhappy, say so before the clock runs out.

The refill, and what happens if it fails

When you approve, two things happen together: the approved amount is released to the freelancer, and your card is charged to top the buffer back up for next week.

If that charge fails, the freelancer is still paid for the hours you approved, out of money already funded. The contract then pauses rather than continuing unfunded, and both of you are told.

This is the right behaviour in both directions: nobody works an unfunded week, and nobody loses money for work already approved. To resume, fix the payment method and reactivate, which funds the buffer again.

Keep a valid card on file. A card expiring quietly is the most common cause of an hourly contract pausing unexpectedly.

Pausing and resuming

Either side can pause. On pausing, the unused buffer is returned to your wallet — you never hold funded time you are not using.

Resuming funds a fresh buffer for the coming week. Pausing between phases of work is completely normal and better than ending and restarting, because the contract, its history and its reviews stay intact.

Ending the contract

Either side can end it. The final week is approved and settled as usual, and the remaining unused buffer is refunded.

There is nothing to negotiate about the unused portion: it was never earned and comes back automatically.

What it costs you

You pay for approved hours, plus the escrow handling fee applied to each weekly funding. Both are shown at funding and appear on the invoice.

Each approved week produces its own invoice, listing the days, the hours approved against each and the rate. A month of hourly work gives you four documents that itemise properly, rather than one lump you cannot break down for your accounts.

When hourly is the wrong choice

If you want price certainty, hourly is not it. The weekly limit caps your exposure per week but there is no total.

If the work has a definable scope, fixed-price milestones give you both certainty and checkpoints. Hourly is for genuine uncertainty, not for avoiding the effort of writing a scope — and using it that way tends to cost more than doing the specification work would have.

Setting the weekly limit sensibly

The limit is the single most consequential number in an hourly contract and it is worth thinking about rather than guessing.

Too low and the freelancer hits the ceiling mid-week and stops, which is disruptive and frustrating for both of you.

Too high and you are committing more money each week than you need to. You only pay for approved hours, so the cost is not real, but the commitment is.

A reasonable approach is to set it slightly above what you actually expect, so a busy week is not blocked, and revisit after a fortnight when you both know what the real rhythm is.

Watching the diary during the week

The diary updates as the freelancer logs, not only at the end. Looking at it mid-week is the single best habit available to a client on an hourly contract.

It costs two minutes and it means you never receive a surprise at approval. If something is going in a direction you did not expect, you find out on Wednesday rather than after the week is billed.

It also lets you redirect. The whole reason to use hourly rather than fixed-price is that you want to steer the work, and steering requires knowing where it is.

When to switch to fixed price

If several weeks in a row look substantially the same, and you can now describe what a good outcome looks like, you have learned enough to scope it.

At that point fixed-price milestones are usually better for both of you: the freelancer can work efficiently without accounting for time, and you get a price. Complete the hourly contract and start a fixed-price one.

The reverse also happens: a fixed-price project that keeps growing is a project that should have been hourly. Ending it and restarting is cheaper than fighting the model.

Comparing the cost against fixed price

Clients frequently want to know which is cheaper. There is no general answer, but there is a useful way to think about it.

Fixed price includes the freelancer's estimate of risk. If the scope is uncertain, they price for the bad case, and if it goes well you have paid for uncertainty that did not materialise.

Hourly removes that premium and moves the risk to you. If it goes well you pay less than a fixed price would have been; if it goes badly you pay more.

So: hourly is usually cheaper when the work is genuinely well understood by both sides and simply not specifiable in advance. Fixed price is usually better value when you want the freelancer to carry the uncertainty and are willing to pay something for that.

What to do when hours look wrong

Before approving fewer hours, ask. Almost always one of these:

The work took longer than expected, which happens and is often nobody's fault.

Time went on something you did not realise was needed — environment setup, understanding existing code, a dependency nobody knew about.

Something was agreed in conversation that you have forgotten or that the freelancer misread as agreed.

The notes are too vague to tell what happened, which is a notes problem rather than an hours problem and is fixed by asking for better ones.

Only after that conversation is a deduction the right response, and by then it usually is not needed.

Common questions

Can I change the hourly rate mid-contract? No. End the contract and start a new one at the right rate.

What if the freelancer logs no hours in a week? Nothing is charged for that week beyond the buffer already funded, which carries forward or is refunded when the contract ends.

Can I see hours before the week closes? Yes, and it is worth doing.

Is there a maximum weekly limit? There is a platform cap, shown when you set the limit.

Related

• The four ways to hire
• Approving work and releasing payment
• Invoices and receipts