Word Import for Confluence

Word documents into a ready page tree — in one drop.

Your company is moving to Confluence, and a hundred and fifty regulations sit in Word files on a shared drive. Each one has to be opened, copied, pasted, then fixed: headings, tables, images. Twenty minutes per document. Nobody was given a week for this, so the move stalls for years.

Getting started

Edit the page, type /Word Import and pick the macro.
Drop .docx files onto the block — or press Add next to a document already attached to the page.
Check what the app recognised, choose how deep to split, press Import.

The app tells you what it found before importing anything — «3 headings · 2 paragraphs · 2 list items · 1 table · 1 image». If that line looks wrong, the import will look wrong too, and you know it in advance.

A page tree, not one long page

An eighty-page regulation is not a page, it is a section. The app splits the document by its headings and builds a tree: a heading 2 becomes a child page of the nearest heading 1 above it. You choose the depth — by heading 1, by headings 1 and 2, or keep everything as a single page.

If a title is taken. Confluence requires page titles to be unique in a space. When the title already exists, the new page gets a number in brackets — «Safety rules (2)». Nothing existing is ever overwritten or deleted.

What survives the import

In Word	In Confluence
Headings 1–6	Page headings, and the split points for the tree
Bold, italic, underline	Same
Lists, including nested ones	Lists
Tables, including merged cells	Tables — merged cells are unrolled so columns stay aligned
Images	Page attachments, placed back into the text
Line breaks inside a cell	Kept

Not imported: footnotes, headers and footers, fields, Word's own table of contents, comments and tracked changes. Only .docx is read — the old .doc, ODT and PDF are not.

Lists come out bulleted

Word keeps the numbering style in a separate part of the file, and telling a numbered list from a bulleted one reliably is not possible without it. We chose bullets: a missing number is easier to spot and fix than a wrong one.

Where your data goes

The document never leaves your computer. It is unpacked and read inside your browser; only the finished pages and their images go to your own Confluence site, using your own permissions. The app has no servers and makes no outbound request of any kind. It carries the Runs on Atlassian badge, which Atlassian verifies at the platform level.

For HR policies, contracts and internal procedures this is usually the deciding argument: the documents you are migrating are exactly the ones you cannot send through a stranger's server.

Known limits

Only .docx.
Lists are imported as bulleted.
Very large documents take a while: everything is done in your browser, and each page is created one by one.
Page titles come from headings; a document without headings becomes a single page named after the file.

Support

Write to support@horizonventures.app. If a document imports badly, send it — we will tell you why, and usually fix it. We answer within one business day.