Add Title and Author in Word's File > Properties. Do not make a decorative first page the only metadata source.
DOCX manuscript preparation
Give the formatter structure, not a fake finished page.
A strong Word manuscript identifies what every passage means. Use document properties, headings, notes, lists, captions, and alt text; let BookSetter resolve the typography and page composition for print and ebook separately.
Propertiestitle and author
Heading 1major elements
Native featuresnotes, lists, captions
No layout hackspagination comes later
Two-minute checklist
A good DOCX is boring to look at and easy to understand.
Apply Heading 1 to parts, chapters, prologue, epilogue, appendices, and named front or back matter.
Use consecutive Heading 2–6 levels for real nested sections, not to obtain a preferred font size.
Create lists, footnotes, endnotes, links, bookmarks, and captions with native features.
Add useful alt text, put captions immediately after images, and retain the original high-resolution files.
Resolve comments and revisions; remove tabs, repeated returns, columns, generated contents, and manual page positioning.
Step-by-step
Prepare the manuscript in seven deliberate passes.
Do each pass across the whole book. Consistency is more valuable than making isolated pages look finished.
- Make a clean copySave a separate import copy. Resolve comments and decide every tracked change.
- Save as DOCXConvert old `.doc` files and create a real document from any `.dotx` template.
- Set metadataAdd Title and Author in document properties and confirm the language.
- Apply structureUse Heading 1 for major elements, Heading 2–6 for subheads, and Normal for prose.
- Use semantic featuresUse native lists, notes, quotes, captions, hyperlinks, bookmarks, and image alt text.
- Remove page hacksDelete tabs, repeated blanks, returns, columns, page furniture, and breaks used only for Word layout.
- Check and confirmRun Manuscript Readiness, repair the source, confirm the Book Map, and proof both PDF and EPUB.
What BookSetter reads
The DOCX importer preserves book semantics.
The readiness check reports what it detected and what required inference. It does not award points for Word fonts, margins, or page decoration that will be replaced.
| Area | Supported evidence | Best input |
|---|---|---|
| Book structure | Title matter, parts, chapters, front/back matter, Heading 1–6, and outline levels | Built-in Word headings with consecutive hierarchy |
| Prose | Bold, italic, underline, strike, small caps, super/subscript, and language | Character formatting applied only where meaningful |
| Complex content | Nested lists, tables, block quotes, figures, captions, links, and bookmarks | Native Word features and consistently named semantic styles |
| Notes | Footnote/endnote references, definitions, and backlink requirements | Word's Insert Footnote and Insert Endnote commands |
| Editorial state | Comments, tracked changes, manual headings, breaks, and missing alt text | A clean copy with every warning deliberately resolved |
| Safety | Package paths, active content, external relationships, archive limits, and XML validity | A non-macro `.docx` from a trusted authoring tool |
Front matter and page numbering
Confirm the Roman-to-Arabic boundary before pagination.
The Book Map assigns every major element to front, main, or back matter. The first confirmed main-matter element establishes the body boundary: preliminary material remains in the Roman regime, that boundary starts the Arabic regime, and later sections cannot switch the book back to Roman numbering.
| Manuscript area | Typical Book Map roles | Numbering behavior |
|---|---|---|
| Preliminary pages | Half title, title page, copyright, contents, dedication, epigraph, foreword, and preface | Front matter before the confirmed body boundary uses the Roman regime. |
| Reading body | Part, chapter, or poem | The first confirmed main-matter element begins the Arabic regime. |
| Supplementary pages | Appendix, endnotes, bibliography, acknowledgments, about the author, and other back matter | Back matter after the body continues the Arabic regime. |
| Ambiguous material | Introduction, chronology, note, or an unstyled heading | BookSetter requires review when evidence is uncertain; the author decides the role before export. |
- Inspect the orderVerify each title page, copyright page, contents page, part, chapter, appendix, and back-matter entry.
- Set the body boundaryConfirm that “Body begins here” identifies the first element classified as main matter, not a later chapter.
- Resolve every uncertaintyIf no chapter or poem is classified as main matter, BookSetter blocks pagination instead of producing an all-Roman book.
- Reconfirm structural editsChanging a role, order, title, start side, or output inclusion invalidates the prior confirmation and requires review again.
Best practices
Tell us what the content is. Do not push it around the page.
| Avoid | Use instead |
|---|---|
| Bold or larger type as a chapter marker | Heading 1 |
| Typed hyphens and manual numbering | Word bullets and numbered lists |
| Tabs and spaces for alignment | Paragraph alignment, lists, tables, or a semantic style |
| Repeated returns to push text down | Let the print composer control vertical rhythm |
| A Word-generated table of contents | Book Map navigation generated for each output |
| Manual note numbers | Native Word footnotes and endnotes |
| Flattened charts with unreadable labels | High-resolution originals, alt text, and captions proofed in both outputs |
AI with boundaries
Let AI explain the evidence, not invent the manuscript.
BookSetter parses the DOCX deterministically. The optional AI preparation assistant receives a structured list of metadata, features, diagnostics, confidence, and repair steps—without manuscript prose by default.
Explain findings, rank repairs, propose explicit Word operations, and prepare a separate copy after uncertain operations are confirmed.
Silently rewrite prose, invent missing text, reinterpret structure without confirmation, or override validation.
Tracked-change outcomes, content edits, heading reinterpretation, and destructive cleanup always remain reviewable. The original file stays untouched.
BookSetter reference file
Learn from the same DOCX the readiness checker can inspect.
The official tutorial is both a complete preparation guide and an import-ready sample. Open its Styles pane to see Title, Subtitle, Heading 1–3, Normal, Quote, Caption, real lists, links, and tables used as ordinary Word semantics.
Keep the downloaded file unchanged as a reference. Duplicate your own manuscript, apply the same semantic approach, and let the readiness report identify what still needs review.
Questions
DOCX preparation answers
Can I upload an old .doc file?
No. Open it in Word, Pages, or LibreOffice and save a clean `.docx` copy.
Will AI rewrite my manuscript?
No. AI may prepare a separate copy from explicit, reviewable formatting operations, but it may not silently rewrite prose or change uncertain structure. The original stays untouched.
Should I format the pages in Word first?
No. Use Word to mark meaning and structure. BookSetter composes print pagination and reflowable ebook behavior separately.
What happens if Word headings do not identify the chapters?
BookSetter blocks pagination when it cannot identify main matter. Correct the Book Map and confirm the first main-matter element before export so front matter uses the Roman regime and the body begins the Arabic regime.
Start with the source, then judge the rendered book.
Upload a clean DOCX, review readiness findings, and confirm the Book Map before formatting.
Guide reviewed July 22, 2026.