VoiceCraft
Menu
Sign in
❦
24.09.2026·By The VoiceCraft team

Getting feedback on a narration before you publish it

Every VoiceCraft book comes with a private share link. How it works, who can hear what, and how to use it to get useful notes from early listeners before you call the book finished.

You have listened to your own book so many times that you can no longer hear it. That is normal, and it is the best reason to let someone else listen before you call the narration finished. A fresh pair of ears catches the mispronounced name, the chapter where the voice sits wrong, the paragraph that runs straight into the next with no breath between them.

VoiceCraft gives every book a share link for exactly this. This article explains what the link does, what a listener sees, and how to turn early listening into notes you can act on.

What the share link is

When you create a book, VoiceCraft generates a share link for it. You find it in the editor, next to the Play all button, labelled Share link. It opens a public, read-only listening page for that book.

A few things are worth knowing before you send it anywhere:

  • Anyone who has the link can listen. They do not need an account, and they do not need to sign in
  • The page shows only chapters that have been narrated. Chapters you have not rendered yet do not appear, so you can share a partly finished book and add to it later
  • Each chapter on the page shows a short excerpt of its text next to the player, so listeners can tell where they are
  • The link does not expire on its own, and there is currently no separate switch to turn it off. The way to take it down is to delete the book from your library, which removes the book, its audio and the share page with it

In other words, treat the link as unlisted rather than secret. It is hard to guess, but anyone you send it to can pass it on. If your manuscript is confidential, for example under a publisher's embargo, share it only with people you would trust with the text itself.

Who to ask

Different listeners hear different things. A small, mixed group usually gives better notes than a large one.

Someone who knows the book. An editor, a co-author or a close reader will catch names and terms that come out wrong, because they know how those words should sound. Our guide to names, numbers and odd words explains how to fix what they find.

Someone who does not. A listener meeting the story for the first time notices when a scene is hard to follow by ear, which a reader who knows the plot never will.

Someone who listens to a lot of audiobooks. They have a feel for pace and for when a voice starts to tire the ear, and they will tell you if a chapter drags.

Two or three people is plenty for a first round.

Ask specific questions

"What did you think?" produces polite answers. Specific questions produce notes you can use. Send the link with two or three of these:

  • Did any word or name sound wrong? Which chapter, roughly where?
  • Was there a point where you lost track of who was speaking or where the scene was?
  • Did any chapter feel noticeably faster, slower or flatter than the others?
  • If the book uses more than one voice, did every switch feel deliberate?
  • Was there anywhere you stopped listening, and why?

Ask for rough locations, not timestamps. "Chapter 7, near the end, the name of the ship" is enough to find the chunk in the editor.

Turning notes into fixes

Rendering in VoiceCraft is per chunk, so most feedback costs very little to act on. You do not re-render the book. You re-render the chunk that needs it.

  • Find the chunk the note refers to in the editor
  • Decide what kind of fix it is: a spelling change for pronunciation, punctuation or a paragraph break for pacing, a different narrator or tier for the voice
  • Edit the text or the settings for that chunk only
  • Press Re-render on that chunk
  • Listen to it together with the chunks before and after, so the fix sits naturally in sequence

The share page picks up the new audio, so your listeners can hear the change on the same link.

Each re-render uses credits, just like the first render. The editor shows the credit cost of each chunk before you press the button, and the pricing page lists the cost per thousand words for each tier. If a round of feedback produces a long list, fix the notes that more than one listener raised first. Those are the ones every future listener would notice too.

When a pattern appears

Sometimes the notes point to something bigger than one chunk. If several listeners say the same narrator tires them over long chapters, try that voice against one or two alternatives on a single chapter before you commit to a change. Our guide to choosing the right narrator voice covers how to compare voices fairly, and working chapter by chapter explains how to trial a change without re-rendering the whole book.

If listeners get lost in scene changes, look at your manuscript before you look at the narration. A blank line where a scene breaks gives the narrator a natural pause, and it changes where chunks begin and end.

Knowing when to stop

Feedback can go on forever. Two rounds is a sensible limit for most books: one with the first complete draft, one after you have made the fixes. After that, notes tend to become matters of taste rather than problems. When the questions above come back with nothing new, the narration is ready.

Before you publish anywhere, run through the self-publishing checklist to make sure the rights and the manuscript are in order.

A short summary

Every book has a share link that plays its narrated chapters to anyone who has it. Send it to two or three chosen listeners with specific questions, fix what they find chunk by chunk, and stop after a round or two. If you need the link gone, deleting the book removes it. And if you have not started a book yet, the voices page is a good place to hear who might read it.