Global Development mid-year status report on 2011/12 annual plan

Summary

Overall, the global development team continues to make progress in building our team; however, we are moving more slowly than would be preferred in some areas. I’m happy that we have made a huge amount of progress in Mobile over the past six months. I would like to be further along in deploying pilot programs in India and Brazil as well as in expanding our grants program. The slower than desired pace is a result of our desire to do a better job of working with the communities where we are deeply engaged, a desire to do more upfront consultation and design work, and due to our relatively thinly spread leadership resource (me). We are also actively reflecting on the Pune Pilot and integrating lessons into how we work across the board, not just in the Global Education Project or in India.

WMF Goal #1: Mobile

Woman taking a mobile picture in Bangalore

We are on track to meet our plan for our mobile target of 2 billion page views for 2011/12 and partnerships with mobile operators representing 500 million subscribers. In December 2011, we had 1.534 billion page views to our mobile sites across all Wikipedias as compared with 802 million in June 2011.^[1] We have made excellent progress across the organization on mobile over the past six months and are in a fundamentally better place than we were. Our mobile partnerships team has built a pipeline of partnerships with mobile operators around the world that start launching in January. Our current partnership list covers key markets in Latin America, Asia, Middle East and Africa, Turkey and Russia representing over 700 million subscribers. Not every deal will come to fruition, but we are confident that some major ones will and we'll begin to attract wider interest in partnerships. In most cases, our partners will be offering Wikipedia access for free to their subscribers and we are working on marketing programs that will expand reach.

The Global Development team works closely with engineering on mobile research, product feature decision making and on technical support for partners. Our engineering team has deployed a much-improved mobile gateway and enhanced its functionality, and is working hard to release an Android app, which closes a hole in our portfolio. They are building out our engineering resources to enable continuous improvement of our mobile position. GD/Eng’s mobile research work (we have done two major studies) has helped inform engineering decisions on the product development pathway. Results from the Mobile Readers Survey 2011 are being analyzed, and will be shared soon. Findings from the mobile research work conducted in India and Brazil can be found at Wikipedia Mobile User Research.

WMF Goal #2: Editor growth

Qatar Convening for Arabic Wikipedia October 2011

Progress on editor growth has been more challenging. We are behind in getting pilot initiatives deployed to really understand the potential for direct impact on editor growth. Our primary effort to date has been the Global Education Program including the Pune Pilot in India. While the program in the US and Canada continues to grow, it has had a small and temporary impact on editor numbers. The program has not been oriented toward creating new Wikipedians, but has added almost 2,000 editors during the Fall 2011 semester, more than thrice the number from Spring 2011 (500+).

The Pune Pilot, which we launched in June, has wrapped up, but was a failure. There were a range of problems involving student plagiarism and the program took on too many students with too few support resources to manage the problems that came up. We also taxed the English Wikipedia community in a way that we had not intended and was regrettable. We learned a lot...and are engaged in a thorough review of the pilot with outside help to ensure we capture the lessons and make better and different mistakes in the future.

We did not have the capacity in place to launch other pilot initiatives in the past six months. We slowed down our plans for Brazil to create space to build a strong relationship with the Brazilian community and conduct some research into the current state of PT:WP. Our India program was at full capacity dealing with the Pune Pilot, supporting the Wikiconference India, and basic program setup requirements. Our India team also took some time to strengthen their links to the community and do a better job of getting early community partnership in program work.

An unplanned for opportunity emerged to accelerate catalyst activities in MENA focused on Arabic Wikipedia. It was not in the annual plan to work in MENA this year, but we took the opportunity presented by the interest of the Qatar Computing Research Institute in supporting Wikipedia. They hosted a small workshop where we met with leading Arabic Wikipedians and laid the groundwork for program work in the beginning of 2012.

Global Development core activity review

Grants and Global South Program

Grants

The Wikimedia Grants program began integrating community review and input through the Grant Advisory Committee in June 2011.
A new program called Wikimedia Participation Grants has been created, co-funded and co-organized with Wikimedia Deutschland, and began operating in November 2011.
Grant approval stats: 23 grants were approved so far; 1 not approved.
Total amount of approved grants: $691,198
- Project/Event grants: $128,187
- Annual Plans (WMAR, WMAT, WMSE, WMHU): $563,011
Prepared revised process and criteria for grantmaking, to be proposed and discussed with the community in Jan-Feb. Revised process to be put in place by March 1st.

Global South Relationships

Participated in and supported community events in Kenya, South Korea and the Philippines
Promoted partnerships between local Wikimedians and local NGOs and government institutions in Kenya, South Korea, and the Philippines
Initiated pilot outreach training effort with ITOCA, a Sub-Saharan Africa information technology outreach NGO. Partnership to include a Wikipedia curriculum in ITOCA trainings, distribution of offline Wikipedia in collaboration with ITOCA, joint development of metrics and evaluation tools for offline efforts
Supported community building efforts by the Wikimedia community in Kazakhstan and its surrounding area

Chapter Relationships

Continued work to ensure all chapters have agreements with WMF and are meeting the compliance requirements of these agreements around reporting
Worked with WMDE, WMUK, WMCH and WMFR to implement the 2011-12 fundraising agreements including the review of their program plans and agreement on fundraising targets
Worked with WMNL, WMAT, WMSE, WMAU and WMHU to develop grants to support their program plans as a replacement for fundraising agreements in 2011-12 and entered grant agreements with WMNL, WMAT, WMSE and WMHU to date
Finalized arrangements with WMIT regarding funds transfers from prior agreement in 2010-11

Mobile

Have negotiated and developed multiple global operator partnerships; a first phase of partnerships have been structured and are ready to deploy in Q1 2012, though we are not yet able to publicize the names.
Worked closely with mobile engineering to design, test and deploy WMF's new mobile gateway and develop new features
Actively involved in the development and testing process for WMF's new Android app including localization in key languages for partner requirements
Worked on new solutions for providing Wikipedia access via SMS and USSD technology. Also conducted a proof-of-concept test with both technologies with an India operator, which will be applied to other partner roll-outs.
Refined and developed framework/templates for working with mobile partners including:
- Trademark Permissions Agreement (in cooperation with legal)
- Mutual NDAs (in cooperation with legal)
- Marketing Plans
- Project Plan Framework for partner rollouts
- Implementation Plan (in development)
Fostered a Mobile Development Community starting at Mumbai hackathon

India Catalyst Program

WikiConference India November 2011

The India Programs initiative has been in launch mode over the past six months. A significant effort has gone into establishing the building blocks. A 4-person consulting team has been put in place - including 2 existing community members (there remains one open slot for a communications consultant). The consultants are all working out of Delhi. Significant time was invested in legal and regulatory work.

The team has been working to build relationships with the Indian Wikimedia community and the Wikimedia India chapter. This has been a learning experience for all involved. The team (mostly Hisham) has involved sharing of the team's plans on Meta seeking feedback, monthly IRC office hours, attendance at meetups, constant lines of communication with individual community members, jointly undertaking activities even if in the background, providing ad hoc support (e.g., helping write grant applications or project design support), building a working relationship with the Chapter and serving as a bridge between community members where inter-community communication & collaboration are problematic.

The consultant team has been engaged in five areas of work:

India Education Program - Pune Pilot. This pilot failed to achieve its goals, created unintended costs for the community, and there are a series of lessons from it that are being documented in our evaluation work. The impact on en-wp community relationships as well as the India Programs credibility has been damaging and both the India consultant team and the Global Education Program team have invested heavily in follow up work to address the issues that arose. Lessons from this inform all the work we are now doing in India, globally in Education and indeed other parts of the Global Development operations.
WikiConference India 2011 was held in November in Mumbai. This was the first ever meeting of the Indian community and had more than 700 people attending (at least parts of the event). The India Programs team provided considerable counsel and hands-on support to the organisers. These included providing 2 Fellowships, providing direction to the organisation, mediating intra-community disputes, motivating the team during the long hard struggle and hands-on work like fund-raising, project review and specific tasks during the conference. The conference provided the Indian community with a positive dose of inspiration.

Hisham Mundol in his Delhi Office

The Indic languages initiative has started off with insightful conversations with dozens of Indic Wikipedians to understand, distill and document experiences, lessons and suggestions for the growth of these communities and projects. Ostensibly, the final intent is to establish pilots but the even larger benefit of this dialogue is to share & cross-pollinate lessons as well as to improve community cohesiveness and build momentum. For instance, the work that we are doing in Hindi is helping to coalesce a relatively inactive community - because right now it's pretty much only a collection of editors.
Improving outreach is a new effort launched in December. The strategic intent is to introduce Wikipedia to new communities of prospective editors. There is very little outreach happening in India and that which is happening is focused on a small geographic area. There is also a degree of skepticism (by some) of the results of outreach and the lack of motivation and capability (of some) to do outreach. This work is focused on a combination of both theory and practice in the sense that there will be a lot of effort put into having outreach tips and standard presentation decks easily available to everyone - but also undertaking outreach efforts directly with community members. Results will be measured and shared to enable and motivate a step-jump in community outreach.
A lot of work has gone into the legal aspects working in India - both in terms of the financial regulatory framework as well as the more complex issues of legal risk around content.

Brazil Catalyst Program

Juliana Bastos Marquis, UNIRIO Wikipedia Ancient History class

Developed a strong working relationship with the Brazilian community via a combination of in-person visits in June, August and October as well as regular two-way communications on mailing lists, wikis, and online chatting
Conducted an analysis of Portuguese Wikipedia to increase transparency on the project's community health and history, identifying the need for more receptive practices on wiki
Secured a commitment from a major computer manufacturer to distribute an offline Wikipedia in all of their computers and inspired the PT:WP community to launch an initiative to improve 5,000 articles to be included in the offline edition
Supported an education pilot at UNIRIO in the improving of Portuguese Wikipedia content
Launched a search for a National Program Director (consultant) which is expected to conclude in Jan/Feb

Arabic Catalyst Program

Organized first Arabic Wikipedia convening in Doha in October bringing together a small, core group of Wikipedians to discuss opportunities to catalyze Arabic Wikipedia
Developed a preliminary plan for the Arabic catalyst program
Conducted initial research and planning discussions for the launch of an education pilot in Cairo
Initiated a relationship with the Qatar Computing Research Institute to support Wikimedia's activities in the Arabic region

Global Education Program

Growth in number of new editors

When it comes to the number of new editors recruited among students, the Global Education Program has exceeded our expectations. At the beginning of this year, we projected a total number of new editors of 2,000 in 2011 (2010: 207) ^[2]. At the end of 2011, we reached a number of almost 2,500 new editors globally.

This growth rate over one year of 1,256%, combined with the fact that we cut the number of staff members working on the Global Education Team by half (Public Policy Initiative: 6 FTE + 1 contractor; Global Education Program: 3 FTE + 1 contractor until November 2011) came at a price: transferring an increasing number of tasks to volunteers has led to a lowering of quality standards that resulted in a growing number of complaints from the existing English Wikipedia editor community. In the U.S. and Canada, our volunteer Regional Ambassadors have not been able to build a sufficient support structure for the high number of students that started to edit in the fall 2011 semester. Some classes did not have any support and some other classes had an unhealthy ratio of Ambassadors to students. As a result, not all students were able to contribute to the English Wikipedia in alignment with our expectations. Also, some of the program's volunteer leaders have taken greater risks than the experienced staff members took in the past (e.g. our volunteer Education Program Advisor Canada accepted a class of 1,700 students at the University of Toronto; this class was supported by 4 Campus Ambassadors which proved to be insufficient). The Global Education Program team thus decided to grow the number of new users on the English Wikipedia in a healthier and more sustainable way beginning in spring 2012 and to put more effort into developing the support structure accordingly. It also embarked on setting standards for the U.S. and Canada Education program (e.g. defining a minimum Ambassadors-to-students ratio^[3]).

Pune pilot

Our first analysis of the Pune pilot indicates that only a small fraction of students in India will be able to contribute to the English Wikipedia in a meaningful way. Interviews with students, professors, Ambassadors and Wikimedia staff members from India make us assume that only few students have the skills (both when it comes to research / writing and English language skills) that are needed to contribute to the English Wikipedia without getting into conflict with the existing community.^[4] Also, only a very small fraction of students in the pilot had a sufficient understanding (if at all) of basic concepts like copyright, plagiarism and academic writing. We should be clear about the fact that the Wikipedia Education Program is not designed to fix the deep problems of the education system in countries like India (where copy-and-paste is tolerated widely and memorizing is the core competency for students who are eager to advance). We can, however, be successful if we limit ourselves to recruiting only those students who already have the required writing and research skills for contributing and teach them how to edit Wikipedia. At the same time, it will be mandatory to do a thorough analysis whenever we fail and adjust quickly to the change that is needed to make the program successful. A post-mortem analysis of the Pune pilot is underway and will be completed in early 2012. Our work in the MENA region already builds on the lessons acquired during the pilot in India.

Copyright violations as a challenge that reaches beyond India

Our first findings from the Pune pilot indicate that copyright violations might be a problem that reaches beyond India. We have anecdotal hints that copyright violations are a bigger problem on other Wikipedia language versions edited by people from the countries of the Global South than we knew so far (we talked to educators who are familiar with the Arabic, Portuguese, and Spanish Wikipedia and who told us that parts of the content in these language versions are copied-and-pasted or plagiarized from copyrighted sources). We will need to be proactive in addressing this issue in our work in these projects and will engage the communities on their own practices in copyvio management.

Community response on the English Wikipedia

In general, we have underestimated the hostility of some members of the English community toward new users who make rookie mistakes. While we generally know that new users on the English Wikipedia are not always treated well, we learned through this term of the Education Program that new users brought on through a Wikimedia Foundation initiative were held to a higher standard than typical newbies. Our students were chastised with more venom than other newbies by some English Wikipedia community members who felt the WMF time would be better spent on improving tools and resources for existing community processes, rather than an initiative to bring in new editors, which puts a natural strain on those existing community processes. It is unclear how we can effectively overcome this problem in the long run. Our experiences on the English Wikipedia also indicate that there is a maximum amount of new users that each Wikipedia language community is willing to accept / able to support. This might affect our plans for editor growth in general, as a too high number of new users within a short amount of time could put too much pressure (or too much perceived pressure) on the existing community.

Growing the program beyond the English Wikipedia

Wikipedia Education Program Arabic

Our first steps in growing the program beyond the English Wikipedia are promising. In early December, we met with members of Wikimedia UK who declared their interest in starting a program at universities in the UK and who are planning to host a Wikipedia and Education conference in the first half of 2012. This current semester, Wikipedians and Wikimedia UK are supporting Timothy Garton Ash, who is using Wikipedia as a teaching tool at Oxford University. In mid-December, Wikimedia Deutschland hosted a train-the-trainer workshop that was led by Global Education Program Director Frank Schulenburg and that got positive feedback from all participants^[5]. Also in December, Annie Lin and Frank Schulenburg traveled to Egypt and laid the foundation for a pilot at Cairo University and Ain Shams University, that will start in February 2012. A pilot in Brazil, starting in spring 2012, will build on the experiences that individual university instructors made in the fall term 2011.

Sharing of best practices and lessons learned

In order to build a truly global community around the Education Program, we encouraged the sharing of knowledge and best practices. We built a repository of sample syllabi in order to encourage the sharing of ideas and lessons. Our new monthly metrics and activities meeting brings people from different parts of the world together. The first three phone conference meetings have reached their goal of sharing experiences and plans for the future with each other.

Revision of the existing training modules

Rod Dunican is working on revising the existing training for volunteers participating in the Education Program. Our goal is to add a training module for professors and to improve the training for Wikipedia Ambassadors based on the feedback that we got in the past. Rod Dunican is working on moving the training online and we expect the first version of this online training to be available in early 2012.

Visual identity for the Wikipedia Education Program

LiAnna Davis and Frank Schulenburg have been working with contractor David Peters on a visual identity for the Wikipedia Education Program. We expect the new visual identity to be available in early 2012.

Communications

WMF Annual Report 2010-11 15dec11

Faster launch, better process, better annual report: With an expanded annual report production team (including for the first time our movement comms manager) Communications was able to assemble and launch a more detailed, translated, and mission-relevant report than all our previous editions. We moved faster and got the report to PDF and wiki publication much earlier than previous years (during a very time and resource constrained period). We'll immediately review the process and focus on a production process next year that i) starts earlier ii) uses simpler design formats, iii) translates completely or at least in more langs, and iv) brings in more community review and input.
Better movement communications: At mid-year we now have a dedicated effort to support WMF staff in their efforts to share mission-critical information throughout our project volunteers, chapters and more. We're increasing our knowledge of simple and fast pre-existing wiki-based systems, and taking more time to propose new systems or hacks to these tools to make them easier to use.
Merch shop makes its debut: Although still a work in progress, WMF now has a merch shop of its own. Taxation, international payment systems, languages and production challenges have necessitated the development of well-designed processes and protocols, but all major pieces are in place for a January community and public launch. Our main vendor/supplier Social Imprints has built capacity to keep up with what we expect will be strong demand, and they respond to our efforts to ensure the WMF and WP brands are handled carefully.
Continued strong coverage of our strategic work: Over the last six months we've seen strong, favorable, and accurate coverage of the Wikimedia's movement goals — largely a follow-up to our surge of coverage from Wikipedia 10. International press are responding well to our calls to action around gender gap, editor recruitment, and our major technology achievements (article feedback, wikilove, and more recently the wiki visual editor).

Offline

Content

Worked with Brazilian community to launch initiative to develop a 5,000 article Portuguese offline Wikipedia for Schools, which entails an editing competition to also improve important Portuguese Wikipedia online
Worked with Indian community to prepare an initiative to develop an Indian specific Offline English Wikipedia for Schools

Tools

Partnered with engineering team to continue improvements to Kiwix interface and begin development of Kiwix for Android (included community dev work at the Mumbai hackathon). In just the first half of the year, monthly downloads of Kiwix increased by 53% (~150% since Dec 2010).
Worked with Google Summer of Code on a "SelectionSifter" to help in article selection for offline content

Distribution

Initiated partnership discussions in Brazil and India for distribution of offline Wikipedia for schools, working with NGOs, governments, and private organizations

Research

Completed a report on mobile qualitative research work conducted in India, Brazil and USA
Conducted Wikimedia's first survey on mobile usage generating important insights for the mobile product strategy
Conducted Wikimedia's first household-based readers survey
Launched the second editor survey in December