[Update 26 Feb 2018: This post was written about Scrivener 2 some years ago. It does not apply to Scrivener 3.]
As a writing tool, I love Scrivener. Unfortunately this comes with some hindrances:

  • Scrivener is not a standard format, so you have to compile and export anything you do to to anything with it.
  • Microsoft Word is a standard format in publishing—obviously people in publishing are a bunch of masochists—but Scrivener’s exports to Word are unstyled.
  • Scrivener’s exports to the .mobi format are barely adequate. Scrivener’s exports to the .epub format are famously incompatible with ebook retailers and need to be cleaned up.
  • Scrivener has no export to InDesign—perhaps to be expected, as InDesign can import Word docs.

What I’m realizing is that I’m going to have to use Scrivener for a first-draft tool only, and then export to Word and finish every work in Word. For someone who loathes Word, that’s a sad prospect.
But what’s more sad is that before I can get going on the Word doc, I have to go through and define all the Styles in the manuscript. And it’s not easy, because the process results in loss of tabs, loss of italicization, loss of any and every special style I defined in my Scrivener manuscript (e.g., styles for text messaging). I have to spend hours going through a novel-length document looking for words and sentences that should be italicized, sentences that should be styled as text messages, and so on.
This has me questioning my entire workflow.
Scrivener » Word (garbage) » Word (manual cleanup) » [final delivery formatting app]
And it’s a huge distraction from writing.
[display-posts post_type=”page” id=”1000″ include_content=”true” include_excerpt=”false” wrapper=”div”]

0 thoughts on “The pain and suffering of Scrivener exports to Word”

  1. Katherine M. Lawrence

    This is exactly what happened to me. I did a lot of manuscript adjustment in Scrivner, only to end up re-exporting to Word.

  2. If your final product is format-heavy, you’re better off doing the final formatting in Word (shudder) or Nisus Writer Pro (much better) — but Scrivener has no problems handling and exporting the basic RTF formats such as italics, paragraph indents, headers, etc. And if you want to preserve certain areas from being overridden during the Compile process, such as your text messages, you can do so with a tool specifically for that. Highlight the text you want preserved, then select Format/Formatting/Preserve Formatting. Scrivener will draw a blue line around the text and exempt it from any formatting processes during Compile.
    Scrivener is not Word and does not approach formatting like Word. It approaches formatting from several directions, depending on what the author’s needs are. The manual is helpful in explaining this.

    1. Thanks, oregon expat, for commenting. Perhaps I wasn’t clear. This issue isn’t formatting per se, but rather application of Styles.
      Styles make the world go around. When using InDesign (or Quark) layout for print, you want to have paragraph and character Styles defining the formatting. This makes it much easier to manipulate later for things like trim size changes, font swaps, etc. When using html and css to format and clean up ebook formats, you want to use classes to define different kinds of elements. Even editors for traditional publishers want manuscripts formatted using Styles—no tabs, no visual hacks. I suppose that that’s because, down the line, that document will be imported into InDesign or Quark.
      Styles are typically semantic. This is a paragraph, this is a blockquote, this is a level 1 header, this is a level 2 header, etc. Scrivener allows you to define styles within the editor, which is all fine and good. In my novel, I can define a particular format for text messages, for example. In my non-fiction, I define formats for the header levels, blockquotes, etc. This is important because these semantics can be preserved as the content moves from medium to medium. A header is always a header, semantically, not just body text made bold and bigger. A blockquote is always a blockquote, semantically, not body text indented and put in a slightly smaller font.
      The problem is that when compiling and exporting from Scrivener, all those styles are lost. For example, in a Word-formatted export, sure, you see the same font, but what you see is not a defined Style, it’s an override of the document’s default normal Style, and when you try to update the document Style to what you want, that update in an instant blows away all of your other emulated text Styles. Why? Because everything in the Word export is Style “normal”; the semantic Styles you used in Scrivener were not included in the export. And that means I have to manually go through the entire manuscript to re-apply all the semantic Styles throughout.
      Sorry to get all geeky and state what might be obvious to you, but I wanted to get across the real problem I’m facing having Scrivener in my writing workflow. If Scrivener doesn’t fix this bug (and yes, I call it a bug), I may need to drop it, and I’d hate to do that. I’ve come to enjoy having the binder and notes right there on screen.

  3. Drat. Just started using Scrivener and bumped up against this limitation while still using the trial. Your post popped up because I was looking for quick n’ easy solutions. There are ways to semi-automate this, using very distinct fonts and character types in Scrivener, so you can search for these in Word and replace with the appropriate style. It’s kludgey, but better than nothing. Not sure if I’m sold yet on Scrivener, overall…


Subscribe to my newsletter

Occasional updates on new posts, what I’m doing, the odd big announcement. Not spammy. Unsubscribe any time.

Scroll to Top