Audiobook narration is quoted in a unit that hides the job. Rates are per finished hour, meaning one hour of completed audio in the delivered book, and producing one takes somewhere between four and eight hours of actual work.
So a rate that sounds like a professional day is frequently a professional day spread across most of a week. Everything else about deciding whether to do this follows from getting that number right.
Where the time goes
The ratio surprises people who have done other voice work, where the recording is most of the job.
Preparation
Reading the book, marking pronunciations, deciding on character voices, researching anything unfamiliar. Skipping this is why narrators end up re-recording chapters.
Recording
Roughly two hours of raw recording for each finished hour, because of retakes, breaths, and stopping to fix a stumble.
Editing
Removing mistakes, breaths and noise, and matching levels. The longest single stage for most people.
Proofing
Listening against the text to catch missed words and mispronunciations. Tedious, unavoidable, and the stage that separates accepted books from rejected ones.
Corrections
Fixing whatever the publisher or author flags, which means going back into the booth and matching the original sound.
The ratio improves considerably with experience — a narrator who has done thirty books is far closer to four hours than eight — but it never approaches one to one, and any plan built on assuming it will is a plan that fails in week two.
The rate is per finished hour. Your life is measured in the other kind.
What the rates actually are
Per finished hour, across the market.
| Tier | Per finished hour | What it reflects |
|---|---|---|
| First titles, royalty share | $0 up front | A share of sales, often small |
| New narrator, flat rate | $50 – $100 | Getting started, building a catalogue |
| Competent, some credits | $100 – $200 | A reliable delivery record |
| Experienced, in demand | $200 – $400 | A catalogue and repeat publishers |
| Established, sought by name | $400+ | Recognition that sells copies |
Work the middle row through honestly. An eight-hour book at $150 per finished hour pays $1,200. If it takes six real hours per finished hour, that is forty-eight hours of work, so about $25 an hour. Reasonable, considerably less than the headline, and entirely typical.
The upper rows are real and they are not reachable quickly. They reflect a catalogue, a reputation with specific publishers, and in the top row a name that moves sales. Getting there takes years and most narrators do not.
Royalty share, and why it is usually wrong
A large share of the work advertised to newcomers offers no payment up front and a percentage of sales instead.
There is a narrow case where it makes sense: an author with a demonstrated audience, existing sales you can verify, and a series where later titles pull earlier ones along. Those authors exist and they mostly do not need to offer royalty share, which is the tension.
What the author risks
Nothing. If the book sells poorly they have lost no money, and they have an audiobook they would otherwise not have.
What the narrator risks
Several weeks of skilled work with no floor. Not a partnership, despite the framing — a partnership implies both parties can lose something.
The reasonable version, when it is available, is a hybrid: a reduced payment up front plus a smaller share. That at least establishes that both sides believe the book will sell.
The room, before the microphone
The equipment order people follow is almost always backwards.
Audio is rejected for room sound far more often than for microphone quality. A reflective room produces a hollow, distant recording that no amount of processing repairs, and a modest microphone in a treated space beats an expensive one in a bare room every time.
After the room: a microphone and interface, headphones, and editing software. The full setup order and what voice work pays across categories covers the equipment side properly, and the priorities are the same here.
Two things specific to long-form. Consistency matters more than in any other voice work, because a listener will notice if chapter nine sounds different from chapter eight — same position, same distance, same settings, ideally the same time of day. And background noise you stop hearing after ten minutes is fully audible across nine hours: fridges, plumbing, traffic, a computer fan.
Auditions, and why they fail
Most auditions are rejected on audio rather than performance, which is encouraging, because audio problems are fixable and taste is not.
The common causes, in rough order: room reflections, background noise, plosives on hard consonants, inconsistent levels, and over-processing that leaves the voice sounding artificial.
After that come the performance reasons. Mispronouncing something in the sample, which reads as not having prepared. A tone that does not match the book — a light comic novel read with gravity, or a thriller read flatly. Character voices that are distinguishable but not sustainable, which an experienced listener can hear you will not maintain for nine hours.
Do not invent a character voice in an audition that you could not still be producing in hour seven.
Read the sample they provide rather than something you already have. It exists to test whether you can do this book, and a polished recording of different material answers a question nobody asked.
Rights, and the clause worth reading
Audiobook contracts are more consequential than most voice work because the recording has a long commercial life.
The central question is whether you are being paid for the recording or licensing it. A flat fee for full ownership is common and is not automatically a bad deal, but it should be priced as what it is: the publisher can sell that audio indefinitely, in any territory, in any format, and you have no further claim.
The other clause worth finding is the correction obligation, discussed above, and the third is the delivery specification. Publishers work to technical standards covering noise floor, peak levels and file structure, and audio that fails them is returned regardless of how good the performance was. Get the specification before recording rather than after, because retrofitting a nine-hour book to a standard you did not know about is close to re-recording it.
Choosing what to accept
Length is the first filter. A thirty-hour epic at a modest rate is a month of your life, and the same month spent on three shorter titles at the same rate produces three credits instead of one.
Genre matters more than people expect. Non-fiction and business titles usually need one consistent voice, which is far less taxing than a cast of twenty. Multi-character fiction pays similarly and costs considerably more in preparation and stamina.
And check what the correction terms are before agreeing. A contract requiring unlimited corrections at no additional payment converts a fixed fee into an open commitment, and the time cost of re-recording scattered lines is disproportionate because matching the original sound is harder than recording it was.
What a first title actually costs you
The figure people quote is the advance or the per-finished-hour rate. The figure that decides whether it was worth it includes what the title consumed.
A ten-hour finished book is somewhere between fifty and seventy hours of work for a first-timer once preparation, recording, pickups and proofing corrections are counted. It also occupies a room in your home for a fortnight, at whatever hours your street is quiet, which is a household cost rather than a personal one and is worth agreeing with anybody you live with before signing.
Against a $200 per finished hour rate that is a genuinely good outcome. Against a royalty share on an unknown title it can be several weeks of evenings for a figure that never arrives, which is the whole reason the split above matters more than the headline rate.
Whether to start now
Synthetic narration has taken the bottom of this market, and it will keep taking more of it. The titles that were previously narrated cheaply because the budget was small are increasingly not narrated by a person at all.
What has not moved is the work where performance is the product: fiction with real characterisation, memoir where the author's tone carries the book, anything where a listener would notice and mind. That has held up, and there is an argument it has become more valuable as the alternative became more common. How synthetic voices reshaped the wider voice market traces the same split across other categories.
The practical reading is that the entry tier is a poor place to plan a long stay and a reasonable place to acquire credits quickly, and that the thing worth building toward is a specific genre and a relationship with publishers who work in it.
Anybody with a language other than English has a materially better position here than in most fields, because catalogues in smaller markets are thin and the competition is a fraction of what it is in English.
For anyone weighing this against other routes, it is contractor work in nearly every case, so what that costs in tax and protections applies, and how the hourly compares against other options is worth doing before committing a month to a first title.