A more efficient way to organize and integrate social news communities



As the number of people using social news websites has increased we've seen huge growth in the number of individual communities being used on each site. While this has been positive in almost every regard, it has also gradually led to the need for a more effective way to organise and browse through communities on each platform.

At the moment, all new communities created on Reddit are simply added to the existing mass; a collection now exceeding 850,000 and growing every day. These can be arranged in order of popularity or searched by keywords but so far that's it. There is no easy way to browse through communities by topic on site and related communities lack any real forms of integration.

The diagram below illustrates this method of organisation:



On a smaller scale, when there were only a few hundred to a few thousand subreddits, this method of organisation was sufficient and an advance from the limited fixed categories of previous systems such as Digg, however, as we now approach the first million its effectiveness is waning.

Users arriving at the site are now faced with an increasingly daunting mass of unorganised communities to navigate. This can make it difficult to find smaller communities and has resulted in a few dominant subreddits receiving the lion's share of the traffic while most remain largely unexplored. Additionally, the lack of integration between related communities encourages reposts as users are forced to post in large numbers of related but seperate communities to reach relevant users.

Fortunately we’ve found that each of these problems can be resolved with a simple but powerful tweak in the websites organisational design. By allowing users to create sub-communities within sub-communities, ad infinitum, and then ranking the collection of sub-communities within each community according to their use, a much more organised 'fractal-tree' model emerges:




Broad subjects are now able to integrate more specific sub-categories, which can in turn include their own further sub-categories, and so on, as far as is useful to users. The wall of a parent ‘branch’ would not only contain posts made directly to it but also posts to all of its various child-branches combined, integrating all their content into a single feed.

Users would be able to navigate their way through these communities starting at the 'root' branches, which would be the broadest, most encompassing subjects, and then narrow the scope of their focus with each further rung of the tree they navigate to.



There are a number of unique advantages to this design:

1. It results in a more intuitive and efficient system to navigate. It's now possible to explore general areas of the site and adjust the scope of your focus as you browse.

2. Communities can better organize their own content. Whether or not a community is interested in integrating with a larger parent-branch, this design would give them the option to create internal sub-categories into which they can organize their own submissions.

3. The total need for reposts in the system is reduced. By integrating related subjects, a post to a specific community such as 'Quantum Mechanics' could, for example, be included within the broader categories of 'Physics' and 'Science' by default. At the moment a post relevant to all three communities would have to be posted separately in each community to reach all their users, resulting in unnecessary duplicates of the same submission.

4. Introducing more 'rungs' or 'layers' into the organizational hierarchy allows for a smoother exposure gradient and a more efficient spread of exposure for rising submissions. In the current social news designs, after reaching the front page of a small niche community it's almost impossible for a post to continue rising to the front page of the total site, even if it's upvoted by everyone that sees it from that point forward. This is due to the large exposure gap it then faces between the front page of that small community and the front page of the main site. This situation is illustrated in the diagram below on the left below. The further the curve falls away from the 'Ideal Reward Gradient' the less likely it is for content to keep rising.



In a system with a nested hierarchy of communities, as illustrated in the image on the right above, however, a popular post to a small sub community would have a much better chance of reaching the front page. After reaching the top of its own wall, it would then begin to climb up the wall of its immediate parent branch where it would receive slightly more exposure and reach a slightly broader range of users. If well received there, it could then continue to climb up the wall of the next parent branch above that, and so on, ascending each rung of the tree until it reaches the front page. The result is a pathway of incrementally broadening exposure for successful posts that enables them to climb even from the smallest and most specific sub communities all the way to the front page of the total site, provided they are met with enough up votes along the way.


Conclusion

In summary, the new organisational structure results in:

  • A more intuitive and efficient system to navigate
  • New choice in scope of focus when browsing
  • Better organization of content within communities
  • Reduced need for reposts
  • Smoother exposure gradients for rising content
  • Better spread of exposure for rising content
  • No loss of previous capabilities
 
Home | About | What We Offer | Walkthrough | Team | Transparency | Funding | FAQs | Contact | BETA