How to turn recordings into a book
Do not start by transcribing. Start by finding out what the material is about across all of it, choose a through-line, transcribe only what serves it, then build chapters from themes rather than from sessions. Transcribing first is the most common way this project stalls.
The sequence, in the order that works
The order matters more than any individual step. Most stalled audio-to-book projects did step four first.
- 1. Inventory: list every recording with date, speaker, subject and length. One spreadsheet.
- 2. Analyse across all of it: what themes recur, where is each stated best, what is repeated.
- 3. Choose the through-line: one claim, question or story the book will serve.
- 4. Only now transcribe — and only the material that serves the through-line.
- 5. Group the transcribed material by theme, not by session or by date.
- 6. Build a chapter outline from those groups. Test it before writing.
- 7. Edit each chapter in passes: filler, then paragraphs, then structure, then voice.
- 8. Write the connective tissue — the material that was never said aloud but the page needs.
- 9. Read a chapter aloud. If it no longer sounds like the speaker, you have over-edited.
- 10. Fill the gaps by recording more. You will find three or four, and they are quick to close.
- 11. Whole-book review before you commit to a final draft.
A separate free preview: Your free private Book Coach
A separate Book Coach preview for shaping the book; it does not transcribe, process recordings or import private audio. A short private conversation, one question at a time. By the end you should have a clearer sense of what your book is actually about, who it is for, the shape it wants to take, and the one next thing to do. It never writes a word for you.
A separate Book Coach preview for shaping the book; it does not transcribe, process recordings or import private audio. It asks questions and gives you your own answers back. It does not write the book. You can copy or download the finished plan straight from the page, free and with no email needed.
Why transcribing first stalls the project
It is the obvious first move and it is the wrong one.
Transcribe sixty hours and you have a 550,000-word document. It is unsearchable in any useful sense, you cannot hold it in your head, and reading it takes a fortnight. Most people open it twice and never again, and the project ends there — not from lack of will but from lack of a handhold.
Work out what the book is first. Then transcription is a targeted job on a fraction of the material, and every transcript you produce has a place to go.
Themes, not chronology
This is the structural move that makes audio-based books work, and it is the one people resist hardest.
Your material arrived in the order it was recorded. That order reflects your interviewing schedule and nothing else — not the shape of a life, not the shape of an argument.
Group by theme instead. All the material about the business, wherever it turned up. All the material about the war, across nine different sessions. Then order those themes by what the reader needs to know when.
For family memoir this is the single biggest improvement available. A life told strictly chronologically starts with a birth nobody remembers and reaches the interesting part on page 120.
Writing the connective tissue
A book built from recordings always has holes, because conversation assumes shared context.
What you will need to add: who people are, the first time they appear. Where and when things happened. How much time passed between one section and the next. The explanation your interviewee did not give because you already knew it. And the frame — why this book, why now, why you.
This is the part that is genuinely written rather than spoken, and it is usually about 20% of the finished book. Budget for it, because authors who assume the recordings are the book are shocked by how much writing remains.
Going back for more
Once the structure exists you will find gaps — three or four places where the book needs something nobody thought to ask about.
This is good news. A targeted second session of forty minutes, with specific questions, is easy, fast, and usually produces the best material in the book, because you now know exactly what you are looking for.
If the person is elderly or unwell, do this sooner rather than later. The gaps you can identify at outline stage are the ones you can still close.
About Your Book Pro
Your Book Pro is a book-writing platform built in Britain. Start with an idea, notes, half a draft, a podcast, or a finished manuscript, and use our purpose-built AI tools for planning, feedback, editing, proofreading, cover designs, and to create print-ready files. Writing is totally free: you can type, upload or speak your book as much as you like, and use only the editing tools you need with Plus, or go unlimited with Pro.
Produce world-class drafts at the fraction of the cost of hiring a human editorial team.
Every account starts with 1,000 welcome credits, and a plain export of your own words is free.
Each guide says plainly when a person does the job better than AI. Guides on legal topics are general information, not legal advice.
The sequence above maps directly onto the tools, and it is the order they are meant to be used in.
Themes from Bulk Audio does step two — reading across everything at once and reporting what the material is actually about, which is the step that cannot be done by listening back. Plan Feedback tests the structure at step six before you commit months to it. Transcription handles step four, producing clean verbatim with speaker labels and checking names against your Book Dictionary. Light Edit does the mechanical part of step seven without rewriting your wording, so the spoken voice survives.
And because it is all one manuscript, the whole-book review at step eleven sees what you actually built rather than a pile of files.
What still belongs to you: choosing the through-line, and writing the connective tissue. Those are the two acts of authorship in an audio-based book, and they are the reason it is yours.
Upload your recordings
See what's in an archive you already have
Email me my free Book Coach linkQuestions people ask
- How many hours of recording do I need?
- Fifteen to forty hours for a full-length life story, which is roughly what a professional ghostwriter gathers. Ten hours makes a substantial family book. Quality of questions matters more than quantity of hours.
- Should I transcribe or summarise first?
- Neither — analyse first. Find the themes across the whole archive, then transcribe what serves the book you have decided to write.
- What if the recordings are poor quality?
- Get the important ones transcribed by a human at £60–£120 an audio hour, and use tools for the rest. Poor audio is where human transcription still clearly earns its fee.
- Can I mix recordings with things I've typed?
- Yes, and most people do. Notes, letters, diaries, half-written chapters and recordings all belong in the same manuscript. There is no requirement to choose one input.
- How long does this take?
- Six to eighteen months for a full-length book alongside other work. The analysis and structure take weeks; the writing and editing take the rest.