Pay
One assessment, two deliverables, one setup
The expensive part is forming the judgement, and it is already done by the time you record.
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
A bundle is only worth offering when the second item costs you much less than the first, and here it genuinely does. Reading the brief, looking properly, deciding what you actually think: that work happens once and then sits there, finished, available to whichever format you output next.
That is the whole economic case, and it sets how generous the discount can be.
Where the minutes actually go
Time a written job honestly and it splits into three unequal parts. Intake and judgement, which is the slow bit. Production, which is typing. Delivery and admin, which is short but never zero.
Add a recorded version afterwards and only production repeats, plus a smaller slice of delivery. For most earners the second deliverable lands somewhere between a third and half of a standalone recording, because the standalone recording carries the judgement cost that the bundle has already paid.
| Sold how | Your time | What you can charge |
|---|---|---|
| Written alone | full intake plus writing | full written price |
| Recorded alone | full intake plus recording | full recorded price |
| Pair, sold together | one intake, two productions | comfortably above the written price, below the sum |
The number in the third row is where the argument is. If the pair costs you roughly 1.4 times a single job and you charge 1.6 times, your effective hourly rate on the bundle is higher than on either item sold alone, and the buyer is looking at what reads as a substantial discount. Both of those things are true at once, which is unusual and worth exploiting.
Record second, always
The written version is the script. Producing the writing forces the judgement into a fixed order, and reading from that order is the difference between a nine-minute recording and forty minutes of restarts.
Recording first and writing up afterwards inverts the saving: you transcribe yourself, discover the structure was wrong, and pay the production cost twice. The premium that recorded work carries is a supply story rather than an effort one, which the post on what recorded formats add works through, and none of it survives if you spend three times as long producing it.
The discount that keeps the card intact
Price the bundle as its own tier with its own name, not as a percentage off two line items. A named tier is a product; a percentage is a precedent, and the reasoning behind that distinction is the whole of the case for discounting by scope rather than by price.
Two rules keep it clean. Never sell the recorded half separately at the bundle's implied unit price, because that turns the bundle into your real rate card overnight. Never let the bundle become the only thing anyone buys - if it is, your written tier is priced too close to it.
What it does to the rest of the listing
A bundle is the cheapest possible way to add a second price point without adding a second skill. It also does the anchoring work of a middle tier for free, since a buyer comparing three things behaves differently from one comparing two.
Buyers, for their part, tend to describe the pair as wanting the recording and keeping the document, which is the same reason they ask judges for something they can re-read later. The recording's pull has outside support: in Schroeder and Epley's 2015 study, employers who heard a pitch rated the speaker's intellect more favourably than those who read the same words. Nothing here applies to output a model can generate on demand: automated scoring already produces both formats instantly, and the tools that package it are free. The bundle sells because one person's judgement is behind both files, which is also what Rate Cock's judge listings charge for when they price a pair.