The hour behind two minutes

A two-minute clip costs the judge far more than two minutes: reading the brief, planning the register, setup, several takes, editing and delivery all surround the recording. What a buyer receives is the compressed output of that longer process, which is the honest reason a review is priced like labour.

Reading the brief, more than once

Before anything is recorded, the brief gets read properly, usually more than once. The first pass is for the request itself; the second is for what is being asked underneath it, since what people actually want is often not what they wrote down. A judge who skips this step and reacts to the first reading tends to produce a clip that answers the wrong question, which costs more in the end than the extra minute would have.

Planning the register and the shape

Next comes deciding how the register will actually be performed for this specific brief: where it opens, what gets addressed first, where the detail the buyer mentioned gets folded in, how it lands. This is closer to a short piece of direction than a mood. A judge working in a demanding register plans this more carefully than a casual viewer would assume, because a register held loosely reads as uncertain on camera. Holding a tone on demand is recognised work: Grandey (2000), in the Journal of Occupational Health Psychology, describes emotional labour as the stress of managing emotions when a work role demands certain expressions be shown to customers.

Setup

Light, angle, background, whatever a scenario calls for if one was requested: all of it gets arranged before recording starts, and none of it is instant. A judge working from home is managing the same variables anyone shooting video manages, on top of everything specific to the brief.

Takes

Very few clips are one take. A line lands wrong, the register slips for a second, the pacing is off, and the judge starts again. Multiple attempts at a couple of minutes each add up quickly, and the version that gets delivered is usually the best of several, not the first.

Editing and review

Even a simple clip gets watched back before it goes out. Trimming a slow start, cutting a stumble, checking the delivery actually addressed what was asked: this is a small step per clip and a real one across a working day of them.

Delivery and the admin around it

Uploading, writing the note that goes with it, confirming it matches what was quoted: this is the part buyers see least and judges do most often, because it happens on every single order regardless of length. For many judges these hours sit alongside other work: Pew Research Center's 2021 survey found 31% of current or recent online gig workers said it had been their main job over the past year.

Why a scenario costs more than a straight read

Everything above applies to a plain response to your material. Add a setting, an outfit, or a scripted beat, and every stage above grows: planning now includes staging a small scene rather than choosing a delivery, setup takes longer because there is more to arrange, and takes multiply because a scene has more places for something to go slightly wrong than a straight-to-camera response does. Props and settings are performance elements with their own scoping questions, and the labour behind them is why a scenario is quoted differently from a straight read rather than treated as the same job with extra flavour.

Why this is the honest basis for price

None of this is a case for a specific number, which is not this site's business and would be wrong within a month anyway. It is the reason pricing is quoted rather than posted on a menu: the visible output is two minutes, the actual labour is considerably more, and a fixed rate has no way to reflect that a scenario with setup and multiple takes is not the same job as a straight read. It is also why length moves the price more than almost anything else a buyer can change in a brief - every extra minute of finished footage multiplies through planning, takes and editing rather than adding once.

Compare this to what an automated score costs to produce, which is close to nothing per request after the model exists - the two are not competing on labour at all, they are different products with different cost structures, and that is most of why one is priced per-request and the other is priced on a card. A number from an automated tool also costs nothing extra to regenerate, which is part of why treating that number as a stable data point works in a way a one-off performance was never built to. If what you actually want is a fact rather than a performance, a measurement taken properly once costs you five minutes and settles the question outright. Weighing the two properly, labour against speed, is exactly what Rate Cock's comparison of a human judge and an automated score sets out to do, and it is worth reading before assuming either is simply the cheaper option.

Seeing the production arc this way is also the fastest route to a cheaper, better quote: a specific brief removes guessing time at the planning stage, which is the one part of this process a buyer has any control over.