Pay
Commission is the least important of the four deductions
A generous cut behind a high withdrawal minimum pays a beginner nothing for two months.
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
Commission is the number platforms compete on, which is exactly why it is the number that tells you least. A platform taking 15 percent with a 100 withdrawal minimum and a six-week hold pays a new earner nothing at all for two months, and a platform taking 25 percent with a 10 minimum pays them in week two.
The comparison worth making is take-home per month in your first quarter, and commission is one of six inputs to it.
The six things to actually compare
Commission. The advertised cut. Note it and move on.
Who pays processing. Card and wallet charges are typically a flat amount plus a percentage. Some platforms absorb them into the commission, some pass them through as a separate line. Two platforms with identical commissions can differ by three points on this alone.
Hold length. The gap between delivery and the money becoming withdrawable. Anywhere from instant to six weeks. This does not change your annual income and it entirely changes your first quarter. The processor underneath can add its own delay on top: Stripe's payout documentation says a first payout is typically scheduled 7-14 days after the first live payment, and can take longer depending on industry, country and risk level.
Withdrawal minimum. The threshold below which you cannot take money out. A high minimum permanently parks a slice of your earnings on the platform, and the smaller your monthly volume the larger that slice is as a proportion.
Payout currency and method. If the buyer pays in one currency, the platform holds another and your bank wants a third, you are paying for two conversions you did not choose, at spreads nobody advertises.
Demand. How many buyers are actually there, and how many people are already serving them.
Why demand outranks all five of the others while you are small
A fee structure operates on money you have earned. Demand decides whether there is any.
Ten percent better terms on four jobs a month is a rounding error. Twice the enquiry volume at worse terms is not. Until you are turning work away, the platform with more buyers wins essentially regardless of what it charges, and the arithmetic only inverts once your calendar is the constraint rather than your visibility.
That crossover matters, and it is the same crossover that decides whether to add more jobs or raise your price. Before it, optimise for enquiries. After it, optimise for margin.
A worked comparison
Assume 12 jobs a month at 25, as an illustration with the assumptions stated rather than a claim about any real platform.
| Platform A | Platform B | |
|---|---|---|
| Commission | 15% | 25% |
| Processing | passed through, 0.30 + 3% | absorbed |
| Hold | 30 days | 3 days |
| Withdrawal minimum | 100 | 10 |
| Monthly gross | 300 | 300 |
| Monthly take-home | 245 | 225 |
| First cash in hand | week 6 | week 2 |
Platform A is 20 a month better and two months slower, and for someone deciding whether this work is viable at all, the second fact is the one that decides it. The trapped float matters too: a 100 minimum against 245 a month means roughly a third of a month's earnings is always sitting somewhere you cannot spend it.
What to do with the checklist
Fill it in before you commit rather than after, because the expensive parts - hold length and withdrawal minimum - are invisible until your first payout and unchangeable afterwards.
Then track the outcome rather than the terms. After thirty jobs your own records will tell you what each platform actually paid per hour worked, which is the only comparison that survives contact with reality, and the records worth keeping for exactly this take about a minute a job.
One caution on spreading yourself. Two half-finished profiles earn less than one finished one, and the reason to eventually hold a second is account risk rather than arbitrage between fee schedules.
What the comparison looks like from the other side
Buyers compare platforms on almost none of this. They compare on who is available and whether the listing looks like it will deliver, and the buyer-side account of choosing someone to commission is a useful corrective to the assumption that they are shopping on price. The reason a platform can charge a commission at all is that it supplies the demand, and the automated tier underneath it - the tools most buyers have already tried - sets the floor your paid listing has to clear. Anything a model produces in two seconds is priced as though a model produced it, which the technical account of automated scoring makes concrete.
For a specific worked set of thresholds, methods and timings rather than a general checklist, Rate Cock's rewards page is where the current figures live.