Pay
Price times capacity, minus the deductions
The ceiling is calculable, and calculating it is more useful than aspiring past it.
Guides on Pay: Pricing, from first principles to the annual review, The ceiling is hours, and it arrives sooner than people expect, What rating work actually pays
Your maximum income from this trade is not a mystery, it is a multiplication. Sustainable jobs per week, times your price, times fifty weeks, minus every deduction between the buyer's card and your account.
Do the multiplication before you make any plan that depends on the answer. The number is usually lower than people expect and knowing it early is worth more than the disappointment costs.
The three inputs, honestly
Sustainable jobs per week is not your best week. It is the number you can produce every week for a year while also handling enquiries, admin, and the weeks when you are ill.
Take your best-ever week and multiply by about two thirds. That correction sounds arbitrary and it consistently beats the number people give when asked directly, because a peak week borrows capacity from the week either side of it. The ceiling on hours is not only about output, either: a 2021 WHO and ILO analysis estimated that working 55 or more hours a week carries a 35% higher risk of stroke than working 35 to 40. The composition of a genuinely full load is set out in what a full-time week looks like, and it contains less billable work than the phrase implies.
Price is your realised average, not your list price. Add up a month of actual fees and divide by jobs. Discounts, bundles and the cheap tier you keep meaning to retire all live inside that number.
Deductions are the part that moves the answer most. Platform commission, payment processing, currency conversion, refunds, and whatever you set aside for tax. How the fees stack up covers the first four, and the set-aside is a habit rather than a calculation.
The calculation
Illustration only, with every assumption stated, because the point is the arithmetic and not the figures.
| Line | Value |
|---|---|
| Sustainable jobs per week | 10 |
| Realised average fee (indexed to 100) | 100 |
| Working weeks per year | 46 |
| Gross | 46,000 |
| Platform commission at 20% | -9,200 |
| Processing and currency at 4% | -1,840 |
| Refunds and disputes at 2% | -920 |
| Take-home before tax | 34,040 |
| Set aside for tax at 25% | -8,510 |
| Spendable | 25,530 |
The gross has lost 44 per cent by the bottom row. That fraction is what surprises people, and it is why comparing platforms on headline commission alone misleads - the argument for comparing on take-home instead is exactly this table run twice.
Note also that six weeks came off the year before anything else did. Holiday, illness and the fortnight after a platform change are not exceptions, they are the normal case.
Which lever actually moves it
Look at the table as a set of multipliers rather than a list.
Raising price moves gross and every deduction proportionally, so a 20 per cent rise is a 20 per cent rise in spendable money. Adding a job per week does the same, and costs you the hours. Cutting a deduction moves only that line.
So the ordering is nearly always: price first, mix second, volume third, fees last. Fees look like the obvious target because they are the visible insult, but shaving two points off processing is worth a fraction of what one price step is worth, and the price step takes an afternoon.
Time per job is the hidden fourth lever and the best one early on. Getting faster raises your hourly rate without touching the ceiling on annual income, which matters if your constraint is hours rather than demand - why your effective hourly rate keeps moving is the fuller version.
When the number is lower than you hoped
This is the common outcome, and there are only four honest responses.
Accept it and size your plans to it. A ceiling that covers a specific thing - a bill, a saving rate, a reduction in other hours - is a perfectly good ceiling.
Change the mix rather than the volume. Move capacity toward whatever your own figures show as the best-paid hour you sell. This is the only change that raises the ceiling without raising either price or hours.
Add a second market. Capacity is the constraint, and a second platform does not add capacity, so this only helps if your slots are currently going unfilled. When a second platform is worth it covers the test.
Decide it is additive rather than primary. Plenty of people reach this number, look at it, and correctly conclude that the work should stay a supplement. That is a result, not a failure, and it is the answer the question of whether this replaces anything usually lands on.
What does not work is planning past the ceiling and hoping. The multiplication does not negotiate.
The parts of the ceiling that are set outside you
Two of the three inputs are partly market variables.
Price is bounded above by what buyers pay for a human assessment when an automated one is free, and the account of what automated scoring reliably delivers is the clearest statement of where that boundary currently sits. Demand for your slots is bounded by how buyers choose among people, which is written up from their side rather than yours in the buyer's account of commissioning a judge.
Precision helps in both directions. If what you sell is exactness, the conventions worth being exact about are documented rather than invented, and a defensible method is one of the few things that holds a price against a free alternative.
Recalculate the table once a year, on the same day you do the rest of your annual figures. Every input in it drifts, and a ceiling computed two years ago is a number about a market that no longer exists. Where a platform's own deductions have changed since you last checked, Rate Cock's judges page is where its current terms are set out, and the commission line is the one most likely to have moved.