Why recorded work prices above written work

Recorded work pays more than written work of the same length, and the gap is wider than the extra half hour justifies. The reason is not that a recording is better; it is that far fewer people are willing to make one.

That distinction matters, because it tells you what you are actually selling and therefore what would erode the premium.

The premium is a multiple, not an amount

Quoting a figure here would be dishonest, because the absolute number depends on your market, your record and your format. The stable part is the ratio.

Across the places this work is sold, an audio review of comparable substance tends to price somewhere around half again what the written version does, and video tends to price at roughly double the written price or a little above. Those are the shapes people converge on, not measurements, and your own card should be built from your own numbers rather than from mine.

The useful way to hold it is: written is the base, audio is base plus a meaningful fraction, video is a multiple. Then check the multiple against your minutes.

The second column moving less than the third is the whole point of the piece.

Why the supply side is thin

Three things keep the number of willing suppliers down, and each is worth naming separately because they have different remedies.

Identification. A voice is recognisable and a face is more so. Anyone protecting a working alias has a genuine reason to decline the format, and that reason does not go away with practice.

Discomfort. Most people find their own recorded voice unpleasant and their early takes stiff. This one does go away, usually within a dozen jobs, which makes it the cheapest barrier to cross if it is the one holding you back.

Fixed production time. A written assessment can be edited in fragments between other things. A recording has to be made in one continuous block, and if it goes wrong at minute nine you are starting again. That is a scheduling cost, not an effort cost, and it is why recorded work resists being squeezed into gaps in a day.

The first of these is a real constraint that deserves a real decision rather than a shrug, and the trade-offs of working under a name versus an alias are worth settling before you list a video tier rather than after your first booking.

Non-copyability is doing quiet work

A written assessment can be forwarded, pasted, reformatted and passed off. A recording of a specific person saying specific things about a specific submission is much harder to detach from its origin, and buyers who care about that are paying for it whether or not they articulate it.

This is also why recorded work is comparatively insulated from automation. Generic written assessment has already been commoditised by anything that can produce paragraphs on demand, and the honest account of what an automated score reliably does is the clearest picture of which part of a written offer is now worth nothing. The wider freelance market shows the same pressure: after ChatGPT's release, Hui, Reshef and Zhou found that writing freelancers on a large online marketplace saw monthly jobs fall 2% and monthly earnings fall 5.2%. Recorded human reaction is not on that list yet, and the parts of it that are - a synthesised voice reading generated text - are recognisably not the product.

What it costs you to supply

Equipment is the smallest part of this, which surprises people who put off starting until they have bought something.

A usable microphone matters, because audio quality is the one technical dimension a buyer notices instantly and holds against you. Everything else is optional. Camera quality is largely irrelevant at the resolutions this work is delivered at, lighting is a window, and editing software you already own is sufficient for cutting a false start off the front.

The real costs are the fixed block of quiet time, the retakes, and the fact that a recorded job cannot be done while somebody else is in the room. Price those, not the microphone. There is a longer treatment of what equipment is genuinely required and it is a shorter list than most people assume.

What would close the gap

The premium survives while supply stays thin, so watch the things that would thicken it.

More earners crossing the discomfort barrier is the obvious one, and it happens gradually rather than suddenly. Platforms making recorded delivery a default rather than an upgrade would do it faster. A crowded category shows this first as fewer bookings at the same price rather than as visible undercutting, which is covered in the piece on early crowding signals.

None of that is imminent, and the direction of travel over the last few years has been the opposite: as written assessment got cheaper to fake, the value of a person visibly doing the work went up. Buyers commissioning a human are largely buying evidence that a human did it, which is a point the buyer-side account of how people choose a reviewer makes better than an earner-side one can.

If you want to see how recorded formats are packaged and sold in practice before you build your own tier, the tools most buyers have already tried set the baseline they are comparing you against, and the way Rate Cock lists judges offering recorded formats shows what a live listing for one looks like.