Which unit you sell in changes what you optimise

Almost all rating work is sold per job, and almost nobody chooses that deliberately. It is worth understanding what the choice does, because the two units reward completely different behaviour.

Per-job

You quote a price for a defined thing. How long it takes is your risk and your reward.

This rewards getting faster, which is the fastest-improving variable in your first six months. A job that took forty minutes and now takes fifteen has tripled your effective hourly rate at an unchanged price, and no negotiation was required.

It punishes unclear scope, brutally. If the boundary of the job is vague, every ambiguity is resolved at your expense, and a single sprawling job can wipe out the margin on three clean ones.

Hourly

You quote a rate and bill the time.

This protects you against scope creep automatically - more work is more hours - and removes the incentive to rush. It also removes the reward for getting faster, which is a real cost given how much faster you will get.

It needs a trust relationship to work at all. A buyer who cannot see the work being done is being asked to accept your account of how long it took, and a new earner has not earned that.

The practical answer

Sell per job, and put the protection in the scope rather than in the unit.

That means naming, in your own listing, what a job includes: the format, the length, the turnaround, how many rounds of anything, and what is a separate charge. Every item you leave unnamed is one you will end up doing free. A structured rate card is the practical form of this.

Keep an hourly figure privately, as a diagnostic rather than a price. Divide each job's fee, after deductions, by its total time including the unbilled parts. Working out the after-deduction figure takes ten minutes and is the number that tells you whether a tier is worth keeping.

The one case for hourly

Open-ended work with a buyer you have worked with repeatedly - an ongoing arrangement, a large comparison across many submissions, something whose end point genuinely cannot be defined in advance.

Here per-job pricing is a guess, and a guess against an undefined scope is a loss most of the time. Bill the time, cap it in advance so the buyer is not exposed either, and confirm the cap in writing.

Buyers usually arrive expecting a fixed price for a defined thing, which is one reason per-job dominates; their side of how a commission normally opens is short and worth reading. Where the job is a scored assessment, defining the unit is easier because the rubric defines it - how those are structured doubles as a scoping tool, and the measurement conventions settle the arguments a vague brief would otherwise start.

The platform this site is published by prices per job by default; the current structure is on Rate Cock's judges page.