Task force/Content scope/Project costs
Quantifying the cost (staff or volunteer) of new projects is very difficult. The actual immediate cost of setting up a new wiki and domain name that receives a small amount of traffic is almost negligible (that's why companies like Wikia can afford to set up many thousands of them). That is, of course, if no new technology is required. To give a counter-example, there's a proposal to set up a Wikimedia wiki using SignWriting, a complex script based on image representations. Reviewing the relevant MediaWiki extension for security, scalability and usability could easily be a multi-month effort locking down a full-time resource. That might be a good investment, but it would be a significant one, even if you apply a low standard for an initial deployment.
Once a project is launched, the costs resulting for the Wikimedia Foundation include:
- Legal costs - e.g. Wikisource running into issues around copyrightability of certain documents it hosts. The risk of legal issues obviously depends on the nature of the project.
- Technology costs - regardless of whether a project requires special software to start, it often will have specific enhancement or configuration requests that need to be managed, starting with small configuration tweaks and going up to a complex code review for some community-developed extension (examples of the latter include the proofreading extensions used by Wikisource, and the DynamicPageList extension used by Wikinews).
- Communication costs - whenever we talk about WMF's activities, both mentioning and not mentioning projects creates different kinds of costs (e.g. people complain when we omit specific or all sister projects, but ensuring that they're always correctly represented takes significant effort).
- Administrative costs - we have to ensure that domain names, trademarks, etc. are properly maintained, that people don't misuse the brand for inappropriate purposes.
- Inbound inquiries costs - any new high-level project can create an anticipation and expectation among third parties that we're interested and able to take on significant partnerships around said project. Those expectations need to be carefully managed.
- General noise - small communities often lack the critical mass to deter trolls and malcontents, and may in fact become a safe haven for people blocked elsewhere (see recent Wikiversity events). This can create noise and distraction on all levels of the organization. Improvements to process and governance may be able to help with this.
On a community level, the costs are even harder to quantify. Clearly, significant community effort is spent maintaining a consistent presence of the "Wikimedia family" in various templates, outreach resources, announcements, etc., and the general noise above often includes extensive and time-consuming community discussions. New projects of limited relevance can be a distraction from arguably more important work, especially when there's an inherent assumption in a "sister project" model that all projects are equally deserving of attention.
Putting it all together, any approved official Wikimedia project creates substantial organizational and community cost. The decision to add to our project family should not be taken lightly.