The problem

Often, when creating a new page about a topic, related pages exist but the author has not thought of looking up (or has not found) related pages and blogposts before starting to write a new page. This has several negative consequences for the wiki content as a whole:

• Existing content is not utilized to save time when writing new pages
• New documentation does not take into account existing documentation
• Lessons are not being learned about projects and processes as much as could be
• Duplicate or conflicting information enters the wiki, which makes searching slower and more difficult
• Old, inaccurate content may cause loss of productivity or even damage business when obsolete processes and instructions are followed
• The volume of content grows faster than it has to and makes managing both the content and the system more time-consuming and resource-intensive

On a related note, sometimes blog posts inadvertently become pages when people make a "hot topic" news item but don't know that a page exists for it or forget to link to it. The blog post is updated 100 times with attachments, comments, tables and so on until it becomes one of the most viewed "pages" of the wiki. And it can't be moved. Or converted. Or made a part of the page hierarchy in any way.

It is difficult to alleviate this problem by simply "educating staff" when time and money is short, new employees keep entering the user base, and people generally have too little time to do proper research before creating a new page.

The solution

When creating a new page, it would be useful to see which pages already exist that relate to the current page somehow. If other pages already exists that describe the topic, the user could review them. Perhaps a new page does not have to created in the first place but existing documentation can be updated instead. Or perhaps parts of the new page don't have to be written at all because they have already been described elsewhere and the user can instead link to existing documentation.

The benefits

The central benefit here is that users become aware of existing, related content and can take it into account better when creating new content. This has several positive implications for the usability and manageability of the system as a whole:

• The overall usability of the wiki soars thanks to better linking, fewer pages, and less conflicting information.
• Management becomes much easier in the long run because there is less content to manage.
• Wikignomes (like me) save hundreds of hours annually in time when they don't have to constantly merge pages, archive pages, link pages, move comments, convert blogs to pages, remind authors to check existing documentation before they create content...

How it works

I'm thinking of a robot much like Hercules, the Atlassian support robot that looks through log files and suggests existing Jira issues based on the error messages found.

The way it would work in my view is that after entering the title for a page and moving the focus away from the title field, Confluence would search for pages that are similarly named or labeled and show related pages in a sidebar on the right. The suggestions would have two purposes:

1. "Similarly named pages exist. Did you mean to edit one of these pages?"
2. "These pages seem to relate to the page you are currently writing. Consider linking to one or several of these pages if they are relevant to the page you are writing."

While I'm at it, this same feature would be awesome in JIRA.