Manuscript Formatting Guide
A few minutes of cleanup before uploading can dramatically improve your analysis quality. FirstReader™ works best when your manuscript follows standard conventions.
Why this matters:FirstReader analyzes your manuscript chapter by chapter. If chapters aren't detected correctly, the tool may treat your entire manuscript as a single chapter, producing weaker findings. The pre-upload preview will show you the detected chapter count so you can verify before paying for analysis.
Supported File Formats
- .docx — Microsoft Word documents. The most common format from writers.
- .pdf — PDF files from any source (Reedsy, Vellum, Word, etc.)
- .epub — EPUB ebooks with chapter structure preserved
Chapter Headings
FirstReader recognizes chapter headings in any of these formats:
- Chapter 1
- Chapter One
- Chapter 1: The Beginning
- Chapter One: The Beginning
- One
- Twenty-Three
- Prologue
- Epilogue
- Part One
In Word documents, you can also use the built-in Heading 1 or Heading 2 styles. FirstReader will detect them and treat them as chapter breaks.
Tip: Each chapter heading should be on its own line, with no other text on that line.
Front Matter
FirstReader automatically detects and strips common front matter, but manually removing these sections produces the cleanest results:
- Title page, copyright page, dedication
- Foreword, preface, introduction (unless it's in-story)
- Table of contents
- List of characters or pronunciation guides (these belong in the character list at upload, not the manuscript)
Back Matter
FirstReader automatically detects common back matter headings, but stripping these manually is the most reliable approach:
- Acknowledgments
- About the author / author bio
- “Also by” lists, sneak peeks, preview chapters of other books
- Discussion questions or reading group guides
- Author's notes (unless they're part of the story)
- Appendix, glossary, or any reference material
Other Formatting Tips
- Accept all tracked changes before uploading. FirstReader will see them as text and may misinterpret revisions as actual content.
- Remove commentsfrom your document. Inline comments are not part of the prose and shouldn't be analyzed.
- Strip page numbers, headers, and footers.These usually don't survive conversion, but if they do, they create noise.
- Use straight quotes or curly quotes consistently. Either works, but mixing them can confuse dialogue parsing.
- Avoid text boxes for chapter titles. Word text boxes are not recognized as part of the document flow.
- Footnotes are not yet supported.If your manuscript uses them (e.g., for worldbuilding notes), strip them before uploading. We're working on a footnote-aware analysis pass for a future version.
Verification at Upload
After you upload, FirstReader shows you the detected chapter count and total word count before you pay. If the numbers look wrong, go back and check your formatting before proceeding.
You can re-upload as many times as you want until the segmentation looks right. We only charge when you commit to an analysis tier.
Need Help?
If you've followed this guide and FirstReader still isn't detecting your chapters or scenes correctly, email us at support@firstreader.appwith a sample of your formatting and we'll help you figure out what's going on.