This year’s Product annual plan builds on the initiatives of the past year with a focus on getting better, learning more about our users, and building on our existing experiences. Our focus is not on grand initiatives, but deep and meaningful improvements in our focus and methods, preparing ourselves to fully support the movement strategy when it is unveiled later in the year. Nonetheless, there are some exciting developments in the works for all of our stakeholders.
We’ll augment the community wishlist with resources focused on programs. We’ll deep dive into supporting smaller editor communities and New Readers. We’ll address platform shifts in editing. We’ll continue to improve our reading experiences to make them richer and more accessible. Finding things on-wiki will be easier and faster, and will have better support. Fundraising tech will continue to support our donors.
The Product team has always been organized around audiences — traditionally Editors, Readers, Discovery, Community Tech and Donors. This year we will enhance this focus, based on more refined and systematic definitions of our audiences as shown in the audience map below. Our initiatives will balance sustaining existing communities with targeted developments focused on new audiences, platforms and technologies.
Our initiatives will be guided not only by deep engagement with our users but by deliberate and thoughtful support of our employees. The Team Practices Group will lead this effort by providing day-to-day support of our teams as well as taking on larger organizational issues. We will also continue to infuse our processes and initiatives with design, and support a greater footprint for this function across the organization.
As we mature, we are moving towards a focus on impact (as opposed to products or technology) using a Program format for our framing. This approach, which was fine-tuned by the New Readers team this year, helps us think outside the "product" — and allows us to work more effectively with additional stakeholders in the Foundation and the Community in support of clear, actionable outcomes. Note that the Product organization is contributing significantly to the cross-org programs detailed in another section, and here we discuss “product department only” programs.
We are deeply moved and inspired by our place in the movement. We believe that a mutual understanding and increasing collaboration with the community will accelerate our common progress towards the sum of all knowledge. It’s an honor to be a part of the movement, and we look forward to 2017–18 with a renewed sense of purpose and hope.
Program 1: Make knowledge more easily discoverable
Wikipedia and the Wikimedia projects have a wealth of knowledge and content that is not easily discoverable by our readers. Discovery’s mission is to make that knowledge more easily discoverable, on all interfaces, to aid in organic growth. Our primary products are Search (includes search querying, geo search, search result recommendations, etc.), Wikidata Query Service, the Wikipedia.org portal, and interactive Maps.
Discovery also maintains the Elasticsearch infrastructure that is used for critical functions such as Logstash and backend search on Phabricator.
Outcome 1: Through incremental Discovery improvements, readers are better able to discover and search for content.
Objective 1: Implement advanced methodologies such as “learning to rank” machine learning techniques and signals to improve search result relevance across language Wikipedias.
Objective 2: Improve support for multiple languages by researching and deploying new language analyzers as they make sense to individual language wikis.
Objective 3: Maintain and expand on power-user search functionality to help with editor workflows.
Milestone 1: Enable more accurate search results based on a combination of methods, signals, and analyzers that we can create within CirrusSearch. The decrease of zero results rate (see also here) will be a good indicator, as well as a corresponding increase in user satisfaction rate.
Milestone 2: Provide more in-depth power user ability within the site and API.
The user experience of search on desktop is old and may significantly impede the ability of users to find content quickly and easily. To combat that, we will improve the user interface of search on desktop to help users find what they’re looking for.
Team: Discovery with Community Liaisons, Wikimedia Deutschland (Objective 2.4)
Strategic priorities: Reach and Knowledge
Timeframe: 9 - 12 months; Outcome 2/Objective 2.4 might take longer than 12 months
Outcome 1: Ensure that the current map functionality is rolled out and supported across the projects.
Objective 1: We are currently investigating various options in supporting the maps project and are sizing our commitments accordingly. At minimum, we will support existing functionality as well as security and operations.
Timeframe: This program is ongoing, but includes maintenance and creation. The foundation needs to invest in the opportunities afforded by modern technology and usage patterns, and bring the Wikipedia experience in line with the expectations of today’s users.
Wikipedia is one of the top-10 Internet destinations worldwide, with one billion devices visiting Wikipedia every month. We provide an important function to hundreds of millions of people throughout the world. However, learning on the site, as approximated by pageviews and number of devices*, appears relatively flat, despite continued growth in Internet usage.
This program is about celebrating Wikipedia’s strengths and leveraging them further. It is about applying technological and design improvements to make it the best encyclopedia possible for our readers.
*We also recognize that the current metric of pageviews and unique devices are not ideal proxies for learning. We believe that time spent and return frequency will add color to our understanding, but consumption is not the ultimate goal: we hope to one day measure our human impact.
Increase the amount of learning that happens on Wikipedia, which we will approximate through pageviews, unique devices (a proxy for users), how often users return, time spent on the site, and other ways.
Through a combination of user research, community discussion, design and product development we hope to make Wikipedia a more rewarding experience for our users, which we believe will allow people to learn more.
Outcome 1: Users have better experiences when interacting with articles across mobile and desktop Web platforms, measured through a combination of metrics focused on user retention, pageviews, and time spent on page (Audience H)
Objective 1: Enable readers to gauge the quality and reliability of an article prior to reading
Milestone 1: Distinguish visually between pages with significant issues and pages that have less significant issues or have none at all
Objective 2: Build on existing work creating better metrics to generate a more nuanced measure of learning
Milestone 1: Begin including at least 1 new KPI relating to knowledge consumption in our reading reports
Objective 3: Increase learning for all users by providing articles that load quickly in various settings
Milestone 1: Integrate at least two new services for article rendering on mobile web
Objective 4: Provide a smoother reading experience by improving the ways that users can flag articles of interest for later consumption
Milestone 1: Improve rendering functionality for PDF creation and book tool
Milestone 2: Publish a plan for providing the ability to save pages across all platforms by assessing current functionality for flagging articles for later reading
Objective 5: Improve the end-user reading experience by applying a cohesive design language across multiple screen sizes
Milestone 1: Through qualitative testing, evaluate the effect of increased responsiveness of the mobile website for larger screens
Outcome 2: High-affinity readers (audience G) benefit more from engaging with our sites, and are better aware of Wikimedia's apps.
Objective 1: Iterate and improve new app features that drive reuse and learning by high-affinity readers
Milestone 1: Develop and deploy a push notifications service for more direct app messaging to users. Deliver at least one type recurring/regular notification for each app platform. Beta test notifications for mobile and/or desktop web.
Objective 2: Run a targeted awareness campaign on wiki based on specific feature use or behavioral indicators that apps can optimize for (iOS)
Milestone 1: Raise awareness of the apps and their specific user value as measured by install volumes
Objective 3: Continue to improve accessibility of the apps
Objective 4: Increase engagement and retention by pushing content to users based on relevance, timeliness, interest, and/or preferences (RI/iOS/Android)
Milestone 1: Convert existing “pull” app notifications to use “push” notification infrastructure
Milestone 2: Maintaining core metrics and positive user feedback and reviews while broadening user base
With Internet access and mobile device usage rapidly growing, many people in our emerging communities access the Internet exclusively using mobile devices. With so many new users coming online, the steep learning curve involved in contributing to our projects leads to high reversion rates early on. We will extend the long-form editing experience so that it is great on all devices as well as investigate other contribution tasks that are appropriate for mobile devices.
Objective 1: Improve and consolidate our unified editing platform so that it's great on all devices [Audience B1]
Objective 2: Invest in our future by researching a radically better multi-lingual content system (both reading and editing), compatible across all devices [Audiences B4, G, H, I]
Objective 3: Collaborate with Reading to create new mechanisms by which to contribute edits with lower context or in small ways on mobile devices [Audiences B1, G]
Objective 4: Maintain editing technologies with very high up-time for all our users. Reduce product and technical debt to modernise our tools and technologies, and to make future changes more effective and efficient [Audience B]
Objective 5: Modernise user interface technologies to encompass mobile and desktop platforms, with continued work on the Wikimedia-wide adoption of a standard Web user interface library [Audiences B, G, I, J]
Objective 6: Support work towards unifying MediaWiki's parser implementations, in liaison with Technology's MediaWiki team [Audience J]
A steady inflow of good-faith new editors, and the tools and resources to shape them into productive community members, is crucial to the long-term sustainability and diversity of Wikimedia projects. However, the number of new editors across Wikimedia projects has declined over the past several years even as the number of existing editors has remained stable. Compared to the experiences and perspectives of existing editors, relatively little is known about the experiences and perspectives of new editors. This lack of understanding is particularly acute outside the largest communities (such as English and German Wikipedias). Tenured users spend much of their time reviewing these incoming edits, many of which are unacceptable because of the simple mistakes new users commonly make.
We will work to improve the experience of our emerging communities, on mid-sized Wikimedia wikis with roughly 500–2,000 monthly active contributors, where we think we can provide the most benefit.
Objective 1: Expand and deepen our understanding of the experiences of new editors in emerging communities based on generative research [Audience B1]
Objective 2: Improve, adjust, or create features geared at the needs identified in the research project. The features will be chosen based on research outcomes, but likely will be aimed at the following:
Providing better contextual guidance to new users facing social or technical obstacles during the editing process (such as a perceived lack of permission to edit, a lack of understanding of citations and linking, or a desire for real-time assistance) [Audiences B1, B2, B7]
Better supporting existing editors in mentoring and nurturing new users (such as by improving communications, helping identify good-faith newcomers, or helping new users express their interests and needs) [Audiences B1, B2, B7]
We are seeking to understand and address our larger wikis communities’ extensive backlogs of work, and provide better monitoring, workflow and communication experiences. Identifying ideal contribution tasks is complex and slow, and monitoring changes has similar issues. Recommending specific contributions for users to make has been shown to increase engagement, when trialled with Content Translation. Some community-developed tools show demand for tasks, like FixMeBot from Labs, and on-wiki portals like the English Wikipedia's backlog page. This lack of support is particularly a key growth inhibitor on established communities (those with more than 2,000 monthly active contributors).
Objective 1: Develop a deeper understanding of how communities share, surface, and select contribution tasks on their wikis by researching community-built systems, especially focused on larger wikis [Audiences B1, B2, B7]
Objective 2: Improve, adjust, or create features geared at the needs identified in our research. The features will be decided upon based on the findings, but will likely be aimed at the following:
Supporting users by surfacing recommendations with community-wide contribution tasks, such as in a dashboard [Audiences B1, B2]
Provide users with recommendations for contribution tasks personalised to their activity and interests [Audiences B1, B2]
Outcome 1: We understand and begin addressing the issues our communities are facing with their processes to collaborate on content and conduct concerns ("workflows") whilst better supporting existing users of structured discussions
Strategic priorities: This program benefits the core contributors in our volunteer communities, making their work on the projects easier and more productive. This supports increased production, curation and moderation of Wikimedia projects as sources of knowledge.
The most active contributors on Wikimedia are the primary creators and caretakers of our projects. As volunteers, contributors have built tools and workflows based on templates, bots and gadgets, with limited support from the Foundation.
Sometimes, these volunteer tools break, and the volunteers who created them don’t have time to fix them. Workflows based on templates and categories add unnecessary burdens on volunteers’ time and patience. Limitations in the MediaWiki software can also prevent volunteers from being as efficient and productive as they want to be.
The Community Tech team works on features and fixes that support Wikimedia’s most active contributors. The core work of the team is determined by the top 10 wishes identified and prioritized in the annual Community Wishlist Survey. The team also spends time working on prioritized wishes from smaller groups of contributors, and from WMF Community Engagement.
Objective 1: In 2017, the team is responsible for investigating and addressing the top 10 wishes from the 2016 survey, including: Global settings, Wikitext editor syntax highlighting, Warning on unsuccessful login attempts, Rewrite of Xtools.
Objective 2: Build features to benefit groups of users that work with WMF Community Engagement, including admins, stewards and grantees, as well as internal tools as prioritized by Community Engagement
Target 1: During calendar year 2017, the top 10 wishes identified on the 2016 Community Wishlist Survey will be addressed. For any wishes that aren’t completed, the team will provide an explanation of the investigation and any decisions. (In 2016, 5 of the 10 wishes were completed, with a sixth wish still in progress.) The team will also address several wishes that benefit smaller groups of contributors, and several from Community Engagement’s priority list.
Currently 50% of all work is planned maintenance or new integrations of payment systems. We have steadily increased the number of payment processor relationships. We have also steadily increased the number of countries where we run campaigns (19% growth last year).
This is the cornerstone of almost all fundraising activities and supersedes strategic priorities. With the current high complexity, there are a number of ways to adjust our work and consider alternatives.
Advancement stores critical donor data in our own installation of CiviCRM. Our instance has many duplicate records in its database. We have developed a method for combining (deduping) records from the same donor. Previously, this had only been done by hand. Our current system did not have a measurable impact on the manual process in our largest campaigns in December. We hope to devote more time to the optimization and expansion of this system this year.
Outcome 1: In FY 16/17 four well-trained team members spent 106 hours on deduping related to the English campaign. One of those employees may not be available for this task next year and a new contractor may help. We still hope to reduce the hours spent. We will need to track hours next year and assess impact.
The Team Practices Group (TPG) is collaboratively building a culture and organization that is well-equipped to focus on problem-solving, effective collaboration, and learning. We support and champion the foundational skills, mindsets, and practices that underlie resilient, healthy, and adaptable teams and organizations. The TPG accomplishes its mission through day-to-day agile coaching and scrum mastering with the Foundation's Product teams. Demand for TPG’s services is higher than ever, and comes from across the entire organization. As a centralized team, TPG brings a useful degree of neutrality and objectivity to working with teams, adding value to the projects by bringing in domain knowledge grounded in context, but without core responsibility for the outputs.
By providing Wikimedia Foundation teams and individuals the support and training to define and solve their own problems, the Team Practices Group improves the organization’s capacity to cope with problems, adapt to change, and leverage opportunities.
Key to this approach is cultivating a safe climate for learning and change.
Through an organizational development and team-practices lens, we take a values-based approach to supporting individuals, teams, and the organization to reach the organization’s highest potential. The TPG accomplishes its mission through day-to-day agile coaching and scrum mastering with the Foundation's Product teams. Over the last year, the TPG experimented with expanding the scope of its operations by creating an “Agile Coach: Organizational Collaboration” role which is focussed on identifying and addressing bottlenecks in order to improve collaboration across all levels of the organization. Work in this domain is proposed to continue in FY 17-18. Demand for TPG’s services is higher than ever, and comes from across the entire organization. As a centralized team, TPG brings a useful degree of neutrality and objectivity to working with teams, adding value to the projects by bringing in domain knowledge grounded in context, but without core responsibility for the outputs.
Wikidata’s first and foremost function is to provide a structured data backbone for the Wikimedia projects. The Wikidata team will focus on two goals in this fiscal year: improving and building out Wikidata; and increasing its reach into the other Wikimedia projects. The primary targets here are Wiktionary, Wikimedia Commons, and Wikipedia.
Outcome 1: Continue to assure the technical and social sustainability of Wikidata (software and community) as the central, structured knowledge data base for the Wikimedia movement and beyond. This will help users detect and fix data quality issues more easily, help institutions contribute in a productive way to Wikidata and are part of the community, and help more users write queries.
Objective 1: Improved constraint reports and related data quality tools: The amount of data in Wikidata is growing. We need to build out existing quality tools like the constraint reports and investigate which other tools will help the editors keep the data quality in Wikidata high without burning them out.
Objective 2: Make it easier to query Wikidata's data with specialized tools and UI improvements: The Wikidata Query Service is an incredibly powerful tool, but t is still hard for the majority of users to use it. A big hurdle is knowledge of SPARQL as a query language. We will experiment with ways to make it easier for people who are not familiar with SPARQL to adapt existing queries and create their own.
Milestone 1: Users can detect and fix data quality issues more easily
Milestone 2: More users write queries
Outcome 2: Continue to increase the reach of Wikidata into the Wikimedia projects. We aim to significantly increase the instances of data uses from Wikidata by Wikimedia projects. We will achieve this by removing barriers for editors on Wikipedia and enabling new kinds of data to be collected within Wikidata and Wikimedia Commons — thereby making Wikidata more useful to the other projects.
Objective 1: Enable automated list generation: A lot of work on Wikipedia is spent on maintaining list articles. This is magnified by the fact that each language edition of Wikipedia maintains its lists independently from the other Wikipedias. With the help of Wikidata we will make it possible for people to create list articles that are automatically updated based on data in Wikidata. This is aimed to especially help small Wikipedias keep lists (like the list of a country’s presidents) updated more easily.
Objective 2: Enable editing of Wikidata's data from Wikipedia via VisualEditor: Editors on Wikipedia care deeply about enabling anyone to edit. With the usage of data from Wikipedia, changes to this data also need to be made on Wikidata, to lower the barrier to participation. We will redesign the input widgets on Wikidata and make them more modular. We will then integrate them into VisualEditor to allow editors to change data from Wikidata right on their Wikipedia when editing an article.
Objective 3: Make it possible to store lexicographical data (Wiktionary support): Lexicographical data can so far not be stored on Wikidata. We will build out the support for lexicographical data in order to support Wiktionary and make this additional type of data available as structured data. This requires new entity types and specialized user-interface pieces for lexicographical data.
Objective 4: Supporting structured data for multimedia files (Wikimedia Commons support): Wikimedia Commons holds a large amount of images. The lack of structured data for the meta data like license, creator, location and more causes a number of significant issues for discoverability, management and re-use of our multimedia content. We will work on one of the big ground-work projects: Multi-Content-Revisions. With this feature, we will make it possible to store both wikitext and structured data in the same page. This will require significant changes to the core of MediaWiki.
Milestone 1: Users can create automated lists based on Wikidata
Milestone 2: Users edit Wikidata directly from sister projects
Milestone 3: First lexicographical data is stored in Wikidata and people query lexicographical data, build tools, etc.