Collaborative paper-writing with non-LaTeX-using colleagues

Excluding LaTeX-focused online collaboration services, such as ShareLaTeX, due to your new collaborators' preferences, I think that you have pretty much two major options, as follows:

• Overleaf (formerly WriteLaTeX)
• Authorea

Both online academic collaboration services support simultaneous use of LaTeX and WYSIWYG rich text mode, Authorea also supports some other formats, such as Markdown and HTML. Both services (to various extent) offer other nice collaborative features, such as data sharing, version control, revision notes and much more. Due to multi-format support, I would prefer Authorea to Overleaf, however, the final decision should be made upon a comprehensive comparison of both services across all available features and your detailed requirements as well as some testing.

P.S. Just for completeness, I will mention two other options. The first is to use blog engines that support both WYSIWYG and LaTeX (most of the major ones do: from WordPress to Jekyll). It's a decent option, but I would prefer one of the above-mentioned dedicated services, for multiple reasons. The second option is to self-host RStudio Server (or maybe a custom Shiny application), which would allow academic collaborative writing, using RMarkdown, but IMHO this is the worst solution possible, as trying to implement various needed features and solve issues, such as version control integration, would bring you and your collaborators a lot of headaches.

If you're willing to drop "online service" requirement you might consider showing them LyX, which is a WYSWIM/WYSWIG editor that:

• Compiles to latex
• Allows one to add basic formatting as in word
• Allows one to add figures, tables, citations as in word.
• Allows one to add raw LaTeX code
• Has super-usable equation editor that compiles code to latex, but you can insert equations without writing latex (you can write it if you want).

Lyx is not so good at sharing --- but I've collaboratively worked on Lyx files using Git repository (you might use any other versioned service), and it was a very good experience.

There is a slight problem --- if your colleagures don't use any version control system (or don't know what it is), in this case it is possible to:

• On windows: use some Git Gui (Tortoise Git). I've seen non-technical people use Tortoise SVN to share documents.
• On Linux: add some scripts that perform VCS operations.
• In comments using Dropbox was suggested --- but Git is obviously better choice in terms of feature-set and stability, Dropbox may be easier to learn.

I worked with someone who didn't use LaTeX, and by the last paper we worked on together, we figured out something that worked for us:

• I wrote it in word, peppered with things like\cite{CiteKey}, \caption{\label{fig_some_figure}This is the figure caption} and \ref{fig_some_figure} (i.e. all the crossreferencing stuff).
• The equations (only a handful) were pasted as images from a compiled LaTeX .pdf.
• The figures were converted to .bmp (word doesn't handle any sensible vector graphics formats, or at least didn't in the version I couldn't be bothered to upgrade from).
• The contents of the relevant .bib file was at the end.

The figures needed commenting on and only the text needed editing so it worked out for us. We used word's "track changes" tools for the text edits. At the end I just pasted the text into a text editor, added about 3 \emph{}s and compiled. It's a complete hack of a way of working but the extra effort was minimal and all on my side, so a saving of effort compared to writing a paper in word, which realistically was the alternative.