Confluence vs SharePoint is usually a question teams inherit rather than decide. Here is a one-question framework for deciding, and a four-step audit for the Microsoft 365 license overlap it can reveal.
Full disclosure before we start: I am Head of Product at ikuTeam, and we build apps for Confluence and Jira. The decision framework below works whether or not you ever install anything of ours. We mention our apps once, at the end, clearly marked.
Most teams never decide where a document should live, Confluence or SharePoint. They inherit the answer. SharePoint exists, so documents go in SharePoint, and Confluence ends up holding links to files that people open in another tab, edit somewhere else, and forget to put back.
That default is worth questioning, because a meaningful share of those documents have no business being in a document management system at all. They belong to a page, a project, a decision. They will be superseded in three months. Nobody will ever audit them.
This article is a framework for telling those documents apart from the ones that genuinely need SharePoint, plus a method for auditing what the current arrangement is costing you.
First, the cost of the current loop
Before the framework, it helps to name what the default arrangement actually costs, because it hides in two different budgets.
The interruption. A page has a budget spreadsheet attached. To change one cell, someone downloads it, opens Excel, edits, saves, returns to Confluence, deletes the old attachment, uploads the new one. Ten minutes for a ten second change, and a copy of the file now lives on a laptop.
You have seen the number for what that detour costs: 23 minutes and 15 seconds to get back on task. It comes from Gloria Mark at UC Irvine, though it surfaced in a 2006 Gallup interview rather than a published paper, so treat it as a good estimate rather than a measurement. McKinsey Global Institute has the firmer figure: knowledge workers lose nearly 20 percent of the week just looking for internal information or the right colleague. Close to a full day, every week, before anyone has edited anything. Asana puts a number on the switching itself, nine apps a day and roughly 60 percent of time spent on "work about work."
None of this research is about Confluence attachments, and we are not going to pretend otherwise. What it establishes is that retrieval and switching cost real time, and a loop that generates both several times a day deserves a look.
The license overlap. The second cost is more concrete. Your organization pays for Confluence. It also pays for Microsoft 365. For most staff that overlap is entirely justified, because they need email, calendar, Teams and the desktop suite. But there is usually a cohort for whom the second license exists almost entirely so they can open a file that already sits on one of your pages. Reviewers who check a number twice a month. Contractors who work in your Confluence but not in your Microsoft tenant. Stakeholders who read documentation and occasionally correct a line.
Nobody audits this, because the two costs sit with two different owners. The overlap is invisible unless someone deliberately looks across both.
Confluence vs SharePoint: one question, two answers
Here is the question worth asking about any document:
Who else needs this file, and for how long?
If the answer is "this team, for the life of this project," the document belongs in Confluence. If it is "the organization, for the next seven years," it belongs in SharePoint. Most files sort themselves cleanly once you ask it out loud.
Confluence alone is enough when: | You genuinely still need SharePoint when: |
|---|
- The document belongs to a specific page, project or decision, and makes no sense outside that context
- Its audience is people who already have Confluence access
- It is a working artifact: meeting notes, drafts, trackers, checklists, a budget still in progress
- It carries no retention schedule, legal hold or records management obligation
- It will be superseded, not archived
| - The file is subject to retention schedules, legal hold or formal records management
- Its audience extends well beyond Atlassian users, including people who will never log into Confluence
- You depend on Microsoft ecosystem tooling for it: sensitivity labels, Purview DLP, Power Automate flows
- It is a large binary, media library, dataset or CAD file
- A regulation or contract names a specific system of record
- Many systems reference it, not just Confluence
|
If your documents land mostly in this second group, the goal is not to move them. It is to stop people leaving Confluence to reach them, which is what a storage connector does: the SharePoint library stays the system of record, and the files surface on the page with their permissions intact. That is a different article, and we build a SharePoint Connector for exactly that if it turns out to be your situation.
The mistake is not choosing one. The mistake is never deciding, so every document ends up in whichever system the author happened to open first. A written rule, even a rough one, beats that.
For the first group, the missing piece is usually editing. If people can read the file on the page but have to download it to change anything, they will keep a copy locally and the page version will drift. Editing in place is what makes the first column viable at all.
Audit your own situation: attachments and license overlap
Four steps, doable in an afternoon, no purchase required.
1. Size the Office content you are carrying. In Confluence, start with CQL:
type = attachment and created >= now("-90d")
If you run Jira as well, the equivalent JQL gives you the other half of the picture:
attachments IS NOT EMPTY AND created >= -90d ORDER BY created DESC
Neither query filters by file type, so sample the results. You are looking for the proportion that are Office documents and PDFs, as opposed to screenshots, logs and diagrams. The Office share is usually smaller than people expect and far more consequential, because those are the files that get edited rather than just viewed.
2. Separate editors from viewers. For a sample of those files, use page history and attachment version history to ask who actually changed each one, and who only read it. You are sizing two groups: real authors, and people who occasionally need to change a line.
3. Cross-reference the license roster. For that second group, ask one blunt question per person: if attachment editing happened inside Confluence and Jira, would they still need a Microsoft license for anything else? For most of your staff the answer is yes, and you should not pretend otherwise. The value sits in the minority where it is no.
4. Size it.
Annual overlap = (users who need editing only) x (monthly seat cost) x 12
Run it honestly. If the number is small, you learned something useful for the price of an afternoon. If it is not, you now have a figure to take to whoever owns that budget, which is a better conversation than "this workflow is annoying."
Where our apps fit
One paragraph, then we stop.
ikuTeam Office for Confluence lets people open and edit Word, Excel, PowerPoint and PDF attachments directly on the page, with version history and real-time collaboration, and without a Microsoft license. There is a Jira and JSM version too. It makes the left column of the framework practical. For the right column, we build a SharePoint Connector for Confluence and Jira/JSM instead, which leaves your libraries exactly where they are and brings them onto the page. Different problem, different tool, and plenty of teams need both.
What we would like to know
We arrived at this framework through customer conversations, which means it is shaped by who happened to talk to us.
If you have drawn this line at your own organization, where did you put it? And did the license overlap turn out to be real once you looked, or did it evaporate because almost everyone needed Microsoft for other reasons anyway? We suspect the answer varies enormously by company size and we would like to be less speculative about it.
