Technical wikis

This article is a notepad for an essay on enterprise and special interest wikis. It's a work in progress, by Matt at Agile.

There are two types of wiki

 * The group or personal notebook, a la OpenWetWare, Tiddlywikis on my laptop, or any number of enterprise wiki spaces in large and small corporations.
 * What Pool & Grudin call 'pedias' &mdash; larger, more ambitious projects to capture encyclopedic, canonical, reference material, a la Wikipedia, or large technical and corporate wiki projects.

Types of wiki audience/community
A community wiki is something between a public wiki, like Wikipedia, and an enterprise wiki. Compared to these two end members, there are some advantages and disadvantages in a community:


 * There's a bigger audience than you'll find in most enterprise wikis. More readers, more editors. This is important, because only a small percetage of the total readership is usually interested in being involved in contributing to a wiki. In Poole & Grudin, there were contributions from 2% of the community after one year. Stocker & Trochtermann reported 8--12% after 2 years. Wikipedia has a contribution rate of [well under 1%], but is a rather special case. My own experience with a large enterprise wiki is a contribution rate of about 5--10% over about a 5 year period. If you're encouraged by these figures, don't be: participation obeys a power law, so only about 5--10% of editors would be considered active. So in a community of about 10,000 people, you can expect 500--1000 people to edit a page, and about 50--100 people to do this regularly. Of those, only about 5 or 10 are likely to be editing in a given week.


 * The audience is still not as big as Wikipedia's. It may, however, be comparable to a sub-community in Wikipedia (those people interested in WikiProject Geology, for example). But even assuming you can get their attention, which takes time, many people would rather contribute to a well-established, well-read, well-supported, well-populated project like Wikipedia, than help build content in a new space. How will you entice people into Yet Another Wiki?

How many wikis do we need?
A professional subsurface scientist in our industry has several wikis at his or her disposal:
 * Wikipedia &mdash; the most comprehensive, but Wikipedia eschews esoteric detail, which is something we want more of
 * PetroWiki &mdash; SPE's port of its REH
 * SEGwiki &mdash; SEG's port of its Encyclopedic Dictionary
 * SubSurfWiki &mdash; this public community wiki
 * AAPG wiki &mdash; a new wiki project
 * Possibly his or her company's wiki
 * Possibly some other wiki I don't know about

It's easy to see that there is a strong case for either combining one or more of these efforts (e.g. PetroWiki, SEGwiki and AAPG wiki, though it's hard to see how this could happen now that two of them already exist), or for paying some attention to ensuring good content flow between them -- or perhaps above them.

Since we want esoteric detail that Wikipedia doesn't like, but since Wikipedia is huge and comprehensive, it also makes sense to figure out how to interface with it. For example -- how can we re-use Wikipedia's content, in accordance with its CC-BY-SA license, when it makes sense? For another example -- how can we help ensure that relevant content in Wikipedia is high quality? The American Psychological Society built a task force to do precisely this for content related to psychology.

Bringing people in

 * People don't really care where they find information and knowledge, though they do pay attention to credibility and relevance.
 * Remove every possible impediment to people finding what they need, and even contributing if they want to.
 * Get out of their way
 * Immediate and personal recognition and welcome

Recommendations
Necessary but insufficient conditions for a successful wiki
 * 500 articles good, 1000 better (Stocker & Tochtermann, Poole & Grudin)
 * Make sure maintence and gardening are someone's reponsibility (Poole & Grudin, Grudin & Poole, Stocker & Tochtermann)
 * Value only recognized after intensive use... so get people in there (Stocker & Tochtermann)
 * Plan for future users from the start
 * Embrace the whole concept of open collaboration (Stocker & Tochtermann)

Things successful communities do
Compiled from Akhavan et al (2005), ACTKM , and inferred from O'Dell & Hubert (2011) , and Shirky (2010).

Failure modes of wikis
These causes of failure of enterprise wikis mostly apply to any small community. From Stocker & Tochtermann (2011).


 * Over-dependence on one person, who leaves
 * Too few content creators
 * Unequal and/or obscure contribution rights
 * Inability to extend beyond tech people... others are intimidated or unaware
 * Insufficient content at start
 * Too many tools
 * Mismatch between management or vision and staff or reality

Governance
"Groups tolerate governance, which is by definition a set of restrictions, only after enough value has accumulated to make the burden worthwhile... The burden of rules has to follow, not lead." &mdash; Clay Shirky, Cognitive Surplus

In a well-defined group or enterprise, conflicts, vandalism (intentional anyway), and rogue edits are rare, even non-existent. Don't try to mitigate problems you don't have!