One account for the trade, from day one

Keep one account that only this work touches, and open it before the first payout arrives. Everything unpleasant about the annual reckoning is caused by mixing, and mixing is a default rather than a decision.

The IRS puts it plainly in Publication 583: "You should keep your business account separate from your personal checking account."

What mixing actually costs

A year of mixed transactions has to be separated eventually, and separating it after the fact means reading several hundred lines and deciding, from memory, which were work.

That memory does not exist. Eleven months later, a payment to a software company is either a subscription for the trade or something you bought for yourself, and the bank statement does not know which. The result is a long afternoon that produces a set of figures you are not confident in, which is the worst combination available.

Separation removes the exercise rather than speeding it up. Everything in the account is work, so there is nothing to classify.

What the separation buys

Four things, and only the first is about tax.

Your annual figures become a statement download rather than a reconstruction, which turns a deadline afternoon into an hour - record keeping that takes one minute is the daily half of the same argument.

Your tax reserve has somewhere to live that you are not carrying a card for, which is the entire mechanism behind setting money aside as you go.

You can see what the trade actually earns, because the balance is not being obscured by groceries.

And you can pay yourself deliberately, on a schedule, rather than withdrawing whatever happens to be there - the distinction paying yourself from the business is built on.

What it costs

Usually nothing, and at most a small monthly charge that is itself a cost of earning.

A second personal current account is free in most countries and takes an afternoon. A dedicated business account costs more, is occasionally required once you register a business, and is worth checking rather than assuming.

The one real consideration in this sector is durability. Account closures happen here more often than in most trades, they are rarely explained and they are never convenient, which is why the choice of provider matters more than the fee - bank accounts that do not close on you is the piece on choosing one and on not keeping your whole float in a single place.

Set your platform payouts to arrive there directly rather than routing them through your personal account first, since a payout that lands in the wrong place is a transaction you now have to explain. Which destinations a platform supports is a platform question, and Rate Cock's rewards and payouts page lists its own.

If a second account is genuinely not possible

Then a separate ledger, and the discipline has to be stricter to compensate.

One spreadsheet, one row per transaction, filled in on the day rather than weekly, with a tag on the corresponding bank line so the two can be matched later. It works, it is more effort than the account, and it fails the moment a fortnight gets busy.

Whichever version you run, keep the records for longer than feels necessary; HMRC, for one, says self-employed records must be kept for at least 5 years after the 31 January submission deadline - retention conventions for this sector are described in Measure My Cock's data material. Keep the tooling minimal while you are at it; a spreadsheet outlasts every app, and the category of tools built around this work is described in Penis Rater's tools material if you want to see what is on offer before deciding you do not need it.

The name on a receiving account can surface in a refund or a dispute record depending on the rail, so the account you separate into is also an identity decision - what buyers typically see of the person they commissioned is covered in Rate Penis's etiquette material, and it is worth checking your own arrangement against it before the first refund rather than after.