Wiki Patterns
grow your wiki
Edit90-9-1 Theory
What is it?
The 90-9-1 theory explains the percentage of a wiki's participation, breaking it down as readers being the highest percent, with minor contributors composing the 9 percent and enthusiastic and active contributors composing 1 percent of the total participants in a wiki.
In his article titled Participation Inequality: Encouraging More Users to Contribute
, Jakob Nielsen explores a phenomenon which affects most online, multi-user communities that rely on users to contribute. Participation Inequality is the tendency for most users to participate very little (if at all) and a few members of the community to account for a disproportionately large amount of the content and activity.
When studied, it was found that user participation generally follows a 90-9-1 Rule:
- 90% of users are "lurkers" (i.e. they read or browse but don't contribute)
- 9% of users contribute from time to time, but other priorities dominate their time
- 1% of users participate very often and account for most of the contributions
Usage
This concept is very applicable to a wiki environment because contribution is fundamental to a wiki's success. While it is impossible to overcome this type of human behaviour, it is possible to change the participation distribution (i.e 80-16-4 where 80% are lurkers, 16% contribute a little and 4% contribute the most). Some ways to equalize participation in a wiki include:
- Making it easier to contribute. Offering a wiki help centre, tutorial information and resources for users can help familiarize users with the environment and allow them to feel more comfortable contributing
- Encouraging editing over creating. For most new users, the thought of a blank white page is frightening. Instead, offer templates and examples which users can reformat to fit their content without having to come up with everything themselves.
- Reward participants. Identify your contributors and reward them using small incentives (i.e. gold stars on personal spaces, or Duke Stars on sun forums).
Example
Some wiki examples show variances in these percentages, but as described above, certain best practices can shift the participant percentages. Here are two examples with percentages.
Wikipedia. The Wikipedia article Who Writes Wikipedia by Aaron Swartz discusses the claims made by Jimbo Wales that on Wikipedia, "over 50% of all the edits are done by just .7% of the users ... 524 people." However, Aaron Swartz describes the story he saw from his studies as this:
"When you put it all together, the story become clear: an outsider makes one edit to add a chunk of information, then insiders make several edits tweaking and reformatting it. In addition, insiders rack up thousands of edits doing things like changing the name of a category across the entire site -- the kind of thing only insiders deeply care about. As a result, insiders account for the vast majority of the edits. But it's the outsiders who provide nearly all of the content."
MSDN Community Content. While not precisely a wiki, the Microsoft Developer Network has Community Content features that are wiki-like. According to the sidebar on http://msdn2.microsoft.com/en-us/library/default.aspx as of December 20, 2007, 1866 edits out of 10851 total edits were made by the top five contributors (three of whom are Microsoft employees). That percentage is slightly above one percent at 1.72%.
| Rate this pattern? |
|
Adoption Patterns
Adoption Anti-Patterns
Wikipatterns.com is sponsored by Atlassian Software, creators of JIRA, Confluence, Bamboo and Crowd
Designed by Headshift Ltd
From my own experience, the 90-9-1 rule seems right and makes sense. I don't agree with the study's conclusion though that you can't overcome the participation inequality - at least in some situations. I think Jakob Nielsen is right on with how to improve it but I think the key one is making participation a side effect. Ideally a wiki can become the primary means of communication within a group rather than just something to do for those that feel so inclined. Once it becomes that embedded part of the process (ie, replacing or at least surpassing email) the accumulation of knowledge in the wiki becomes more of a side-effect of the communication that's already occurring.
This pattern could become more useful if re-written according to the pattern template. When I can take the time I will most likely do it, but if someone else has time soon, go ahead, don't wait for me.
With a little bit fear of falling into the [OverAnalyzer] pattern, I suggest defining as part of this pattern, a criteria to measure the level of collaboration in a wiki and to find out where you are in the 90-9-1 theory.
I think this would give advocates/evangelists more ammunition when defending the use of wiki tools in an organization by providing a tangible value (since we first rolled out X wiki tool, we have seen a growth of 40% of collaboration type of thing).
Once a common criteria is defined, we can compare apples with apples when discussing collaboration. The criteria could easily be implemented by the different tools.
Makes more sense in a corporate level, when can account for all the users.
The criteria should consider the following parameters
Just to get the discussion going... or maybe there is something already there...
Ben Gardner is persuasive in his blog post
that "wikis inside the firewall can significantly beat the 90-9-1 rule."
> Ben Gardner is persuasive in his blog post that "wikis inside the firewall can significantly beat the 90-9-1 rule."
"Ben Gardner suggests" or "Ben Gardner relates a story", but he certainly doesn't give any argument, "persuasive" or otherwise.
You might consider rewarding contributors by installing some kind of counter that counts posts and comments and then allow participants after x number of posts to post a link to their specialized information site. This would allow participants to monetize their efforts much like some forums do.
Is this pattern for a specific type of wiki? For example, some of the collaborative wikis for project management, knowledge management seem to have a different ration of readers to contributors.
It seems a very consistent pattern for a wiki where users arent compelled to update. We have a wiki on an intranet for collaboration and information sharing. That shows after a month or two of usage with 900 users set up with edit access:
From my experience, the 90-9-1 theory defiantly holdsfor initial setups of Wikis. I've setup a wiki in an Enterprise, and it was the case for several months. However we are now seeing over 30% contribution rates in a user base of over 900. I've blogged about the Wiki adoption techniques
that worked for me if anyone is interested.
Seems like there's probably a zero missing somewhere in this math .... ?
Tony's initial comment rings true with me.
I think that is what this idea is all about: http://stackoverflow.com/
In my experience these sort of thing are more likely to follow the Pareto Principle. This was created by an economist called Vilfredo Pareto when he noticed that there was a wealth related pattern that was generally present that 80% of the wealth of a country was owned by 20% of the people. When he and others started to look they found that this rule applies to many other phenomena and situations.
A common one that is relevant to the creation of a Wiki will be that 80% of the Wiki created is created by 20% of the users.
I´ve translated the 90-9-1 pattern to spanish in Distribución desigual de la participación en blogs, comunidades, redes sociales, etc...: 90-9-1, algunas claves
Best wishes.
Building out on LeapingWolf view of Pareto principle and the application to Wiki. This principle is the principle of Human behaviour and can be seen time and time again in all social forums. The distribution of participation is a power-law distribution. Ever been in a sporting club where a small number of people do all the work and a large number have limited participation - human behaviour. Why do the larger number have limited participation? Understanding this norm of human behaviour will help us understand how to increase participation in our Wiki.
It would be interesting if wikis on the WWW and Enterprise Wikis both follow this pattern. In a company with - let´s say with 500 employees only 5 would add new articles?
Fascinating site this is. We've been running our own wiki (on our own software) since 07, but had never come across wikipatterns before! Anyway, this seems about the same ratio we're getting. We've been working hard to incentivize especially that 9% group to become bigger, but it's certainly no easy feat. Anyone have any ideas on what kind of incentives work (we've got monetary in place and currently working on a 'hit xxx characters and we'll donate xxx to charity' project). But I'd love to hear other ideas!