Grants talk:APG/Proposals/2015-2016 round 2/Wikimédia France/Staff proposal assessment
Wikimédia France feedback
First, thank you for your assessment, we know they require a lot of time and work to you too. We hoped to publish our feedbacks and reply before the recommendation, unfortunately the schedule given to us this year was so tight that we were unable to do it.
However, you will find below our replies and arguments where we felt it was significant to clarify.
We are aware that it won't change the recommendation which is done by the FDC, but we wanted to share our replies drafted before, not only to respect the time you have dedicated to our proposal and to show you that we have deeply read and understood it, but also to share our different point of view when there are.
About the active editors, we didn’t count the WikiMOOC editors (data were available one week after we submitted our HalfYear report). So we plan to include these data for next year with the WikiMOOC.
For now, numbers for the WikiMOOC are :
- 6000 people registered to the MOOC
- Active editors : 36
- Individuals involved : 960
- Bytes added : 2.979.128
- Articles created and/or improved : 3.074
- 878 new accounts (but only 24 newly registered related to this global metric due to the definition “New accounts created is an editor who has created a new account as a result of a project up to two weeks before an event.” About the MOOC, the first week was dedicated to the creation of an account so attendees didn’t create an account before the project but during.
About the new and improved articles, we have a new and significant objective of 25,000 articles creation / improvement through Ma Commune Wikipédia. The rest is about articles which will be improved thanks to contests supported by Wikimedia France (e.g. re-use of pictures in articles). Such data were not included in our previous reports.
In the FDC application and Impact Report for the previous year, our targets and indicators corresponded to a quality approach we had in place before the global metrics were implemented. We wanted to raise awareness and involve various stakeholders (employees, volunteers, partners ...) to enable them to take into account the assessment and measurement of the impact of our actions.
The “Global metrics” (GM) have been imposed on us when we submitted our FDC application in April 2015 and the deadlines were too short then to include objectives related to these indicators. The impact report we submitted a few months later was an opportunity for us to both implement the GM and integrate additional indicators from our former quality process, thus the overall metrics were applied to the key-actions of 2014-2015.
To define and control the definition, scope, evaluation tools of the GM, the employee in charge of evaluation within Wikimedia France took the online course to master WikiMetrics and got involved in a constant discussion with the team in charge of developing the tool. It allowed him to also train other chapters’ employees on this tool.
Following various chapters’ half year reports submissions, we compiled the data to see how our organization was positioned from a quantitative standpoint :
|Wikimédia Nederland||648||45||3.000 to 800.000||7.600||1279||59.000.000.000|
|Wikimédia Osterreich||632||162||9.128||2.449||6.853||1.000.000.000 max|
We also read the explanations included in other chapters’ GM tables, and we saw that we do not all go by the same definition, when it comes to GM, e.g.:
1. # of active editors involved
Wikimedia Armenia :
This is an estimate number, which involves active editors on Wikipedia, Wiktionary, Wikisource and Commons
2. # of newly registered users :
Wikimedia IT :
We track this metric as part of general data collection but we do not consider it a goal in itself. In fact, in the comparisons within each program, we found the number of newly registered users to be inversely proportional to the actual success of an event, because the more users register the less engagement we can afford for each of them. The largest part of this figure (2k) comes from a CTA via Google Grants which directed Google Search users to an Italian Wikipedia registration form with relevant context. We depend on WMF for exact data; this is the most conservative estimate in a 0–27k range, based on secondary actions.
3. # of individuals involved
For the communication program, the largest contribution came from a long presentation-interview given by the WMIT president on a cultural TV channel: we considered the official average audience figure for the channel (Auditel) as a conservative estimate of the number of people who managed to receive as much Wikimedia information as they would have with a Wikimedia presentation at a conference.
we targeted 'by name' at least 3000 people; via national media we will have reached hundreds of thousands more. The talkshow RTL late night in which two of our volunteers appeared regularly draws 800.000 viewers.
6. Absolute value of bytes added to or deleted from Wikimedia projects
Added value cannot be measured in terms of bytes, as the presence of bot wranglers and photographers immediately skews the results. Some of our top contributors are only active on Commons, yet their contributions should not be counted. As much as we would like to believe it, not all edits counted by the Wikimetrics tool are directly connected to our programs. There simply is not enough granularity. As a result, counting actual bytes "moved" borders on the tedious and/or meaningless.
Please note: WMDE does not track this metric for two reasons: 1) the usability of the Wikimetrics tool is inadequate, making it too time-intensive, especially given a large amount of activities, and 2) the tool would require tracking of user names to aggregate ‘bytes added’. We do not practice this with regard to EU privacy laws and privacy preferences of the German-speaking editing community.
This may explain why the results of our Half-year report are lower than for other chapters:
- We count data only from actions led by Wikimédia France (meaning actions proposed by either volunteers, partners or staff members). We do not count actions of French-speaking editors who are not taking part in our actions.
- From our standpoint, “participants” to an action are people involved in the action. We do not count, for instance, the audience of a TV program in which Wikimédia France is mentioned.
- We are very cautious when it comes to respecting the guidelines provided to count the number of bytes added, and we do exclude the namespace corresponding to media files. Otherwise, the results would be totally biaised.
Feedbacks on questions and concerns regarding the exact definition of the GM have been made several times and WMF itself asked for it through a consultation a few months ago.
Beyond this, it is also important to keep in mind that WMFR has several strategic lines which encompass its actions. Increasing content is one of them, but is equally important as supporting volunteers locally, communicating properly, managing international actions and getting involved in other French-speaking areas, designing new tools and funding our own organization. The typical model actions which provide results on all GM are edit-a-thon on the one hand and photo contests on the other hand. We strongly support theses actions, while developping other initiatives too.
Our chapter is a national organization and the logic of the FDC process is relevant in this regard. Wikifranca, an informal organization at the moment, promotes and supports the Francophonie within the movement, and therefore has developed a number of initiatives which we work on or initiate. So far we have sought a balance between the territorial aspect and the language aspect. When it comes to the latter, and more specifically the organization of the Francophone wikiconvention, other issues are taken into account, related to the Francophonie.
Indeed, there is a geographical bias between North and South on the Wikimedia projects. It's not new and it is one of the priority issues on which our movement is mobilized, so that there is more content available documenting these territories, by editors who live there. Today, more than 90% of contributors on Francophone Wikimedia projects, are from France, Switzerland, Belgium, Canada meaning that 5% of Francophone countries contribute more than 95% of countries recognized as belonging to the Francophonie. Wikimédia France, is one of the chapters, within the Francophonie, which is the most active in developing accessibility to Wikimedia projects in francophone Africa in particular. It plays a central role in supporting organization of the convention, so that it benefits all francophones in our movement.
Thanks to grants from other partners external to WMF, we will welcome at this event over 15 people from Francophone Africa. They have a very active role in their country, where groups of users begin to federate and to carry out various actions.
Support to photographers
The direct aim of this programme (low cost : less than € 3,000 and 50 hours worked) is to directly support Wikimedia Commons editors, so that they can focus on contents of interest for them, and increase their motovation to contribute to Commons.
Open Data Symposium
Our aim is both to position ourselves visibly as an important player of the open data in France as well as educate librarians, archivists, etc . to Wikidata : beyond simply providing content, it is very much about making quality content providing as easy as possible, and getting major stakeholders on board.
Comparisons of amount
First, if one wishes to compare the amount of our grant request this year to another amount of money, it should be - in our opinion - a comparison to the amount of our grant request from last year, not the one we actually obtained. In this case, it increased only by 6 %, not by 18% (from € 600,000 to € 636,000).
Two elements explain this increase of 6 % compared to the grant requested last year:
On the one hand, our fundraising campaign was problematic (this was explained in our Half-year report). This issue could potentially be solved in the coming weeks, but we cannot take such a risk. This uncertainty led to major reassignments for the entire team, over several months, and resulted in disorganization interfering with our planned activity.
On the other hand , the French labor law provides that each employee is increased regularly, at least to compensate the cost of living. Even if we were planning no recruitment, payroll would necessarily increase. In addition, during the WikiMOOC design process led by Jules78120 we found that such a profile could be an extremely valuable resource, and a rare one. Jules is an editor (an admin on French Wikipedia) extremely well known by on-line communities, his involvement is acknowledged by everyone, and he’s especially famous for his tact and impartiality. Moreover, Jules is a journalist, which gives him a true skill to explain and articulate things.
Therefore, in addition to our two staff members dedicated to communities management throughout France (for off-line communities), we would like to hire this third one for on-line communities.
To answer another point on the revenue structure, it should be noted that within WMFr, the fundraising is - at its best - allocated with 0.5 FTE.
We made the decision to allocate as much workforce as we could to the central functions of the association (the link with communities). The time spent on this search for funding, apart from maintaining the relationship with our donors, mainly focused on one-off grant applications for specific actions. To significantly increase our financial independence, we will necessarily need to strengthen human resources.
This being said, the upstream work done for labels and certifications offers a real guarantee of reliability in the eyes of our financial supports and should provide us with external funding opportunities.
About the recommendation
Impact cannot be limited only to Global Metrics results, as argued just above, especially since they are not homogeneously reported thoughout our movement.
The impact on our movement (regarding working practices, documenting, community engaging, actions implementation, etc.) is increasing proportionally to the means deployed, by the FDC in particular.
From our standpoint, it is very important to consider all faces of impact and not to shrink it down to one specific field - which, moreover, still needs to be sharpened.
As indicated, we are going to review our action plan in the coming weeks, to highlight clear strategic priorities. This should meet the expectations from the staff assessment.