Is an audiobook worth it?

Two questions decide it

The audiobook question arrives late. The manuscript is finished, the cover is approved, and someone asks whether you want an audio edition. A quote appears, usually somewhere between two and five thousand dollars.

Two questions settle the answer.

Do the people who buy from you listen to things, or do they read them? And is your voice part of what they are buying?

When both answers are yes, an audiobook is one of the better small investments in the whole project. When either is no, the money is better spent elsewhere, and the rest of this explains why.

The market is growing, and mostly not for your book

Start by taking the growth story out of the decision, because it will be quoted at you.

The Audio Publishers Association runs an annual survey of the market. Audiobook revenue in the United States reached $2.43 billion in 2025, up 9% on the year. That is a healthy market and it keeps expanding.

Two details matter more than the headline. Fiction takes about two thirds of that revenue, and general nonfiction takes roughly a tenth of it. And the number of active titles grew 43% in a single year, to more than 750,000. Supply is growing faster than demand.

So the growth is not a reason to make your audiobook. Your buyers are the reason, or there is no reason.

The right reason: your buyers already listen to you

Think about how the people who hire you encounter you now.

They meet you on podcasts, in the car, on a treadmill, or through a conference talk they listened to twice. That buyer takes in expertise by ear. For them an audiobook is the same relationship continued at length, reaching them during hours when reading is impossible.

The pattern is common among the buyers experts want. Senior people with long commutes and no evenings listen to more books than they read. If your last three clients found you through a show, your book has an audio audience waiting.

If your buyers found you through search, a referral, or a document that crossed their desk, the case is weaker. Make the audiobook later, or not at all.

Reading it yourself is the whole point

Here is the part the production companies underplay. For an authority book, your own narration is most of the value.

A listener who heard you on a show hears the same person. The same pauses, and the same dry aside when a story turns. They already decided they liked how you think. Forty thousand words in your own voice deepens that, and a stranger reading your sentences politely does not.

Now the catch. Narrating is work, and it is harder than talking. Professional narrators produce roughly one finished hour for every two hours in the booth. Published production guides put total labor, including editing and mastering, at around six hours per finished hour. A seven-hour book means several full days of recording, plus the retakes when a plane goes over.

If you dislike the process, it will be audible. Some experts should hire a narrator and accept a smaller version of the benefit. Others should skip audio and spend the money on copies to mail.

What it costs, and what Audible pays now

Narration is priced per finished hour of edited audio. Rates in 2026 run from about $150 to $400 per finished hour for capable professionals. Union and premium talent sit above that. A seven-hour book at $300 comes to roughly $2,100 for narration, before editing, mastering, and cover art.

Narrating it yourself removes the narration fee and replaces it with your days. You still want an engineer to edit and master.

There is also a no-money-down option that deserves its plain name. A royalty-share deal pays your narrator half your royalties for seven years, exclusive to one retailer. If the book sells at all, that costs several times the flat fee. It is a loan rather than a gift.

The payout side changed this year, and it changes the arithmetic. On May 26, 2026, Audible opened a new royalty model to all creators through ACX. Rates rose to 50% for exclusive distribution and 30% for non-exclusive. Membership listening now pays from a pooled model rather than a fixed share of each sale. The legacy 40% and 25% rates are set to end on December 31, 2026. Existing titles have to be enrolled in the new model by then or removed from distribution.

For my own clients this sits inside the ongoing engine. The system prepares the script, the retailer setup, and the metadata, and I handle the uploads and the proofs. The narration is the one part that stays with the author, which is the point of it.

Rough arithmetic on those rates. A $20 audiobook sold outright under exclusive distribution returns about $10. A $3,000 production therefore needs on the order of 300 sales to cover itself. Membership listening pays differently and is harder to forecast. That uncertainty is itself a reason to treat audio as a relationship asset rather than a revenue line.

Machine narration, and why it does not solve this problem

Synthetic narration is now free and fast. Amazon's Kindle publishing platform offers a virtual voice option for eligible ebooks. It produces an audiobook in under an hour. Independent tools do the same for a small flat fee. The voices are far better than they were two years ago.

Listeners have not followed. The Audio Publishers Association's June 2026 survey put machine-narrated titles at 0.03% of audiobook sales in 2025. Stated willingness to try one fell from 70% in 2025 to 61% in 2026. Only 16% of listeners said they had heard one.

Read that against your own reason for making an audiobook. If the point is your voice, a synthetic voice removes the point and keeps the cost of the retail listing.

Machine narration earns its place on a workbook or a reference title. It also suits a backlist book whose audio edition would otherwise not exist. For the book carrying your authority, it is the wrong tool.

When to skip it

Some books should not have an audio edition, and saying so saves people money.

Skip it when your book leans on frameworks, tables, exercises, or diagrams, because those passages punish a listener. Skip it when your buyers read rather than listen. Skip it when you have neither the days to narrate nor the budget to hire well. A cheap human recording sounds worse than no recording at all.

And skip it for now when the book is new and quiet. Audio rights are yours to use later, which is one reason the publishing path you choose matters. A book that finds its audience in print can get its audio edition in year two.

The answer

An audiobook is worth making when your buyers listen and when you are the one reading it. Both conditions, not either one.

Met together, a few thousand dollars and a week of your voice reach a buyer when reading is impossible. Forty thousand words, in a voice almost nobody can imitate. Met halfway, you have a file on a retail page. Met not at all, you have a quieter copy of a book you already own.