Wikimedia Audiences/2016 Product Summary

From MediaWiki.org
Jump to: navigation, search

Last year we built a 2016-17 Annual Plan for the first time based on an FDC format.  As we evolve this effort and move into 2017-18 planning, we thought we’d review and highlight some of the past year’s efforts for Product in context of the Annual plan.  We list out items associated by our strategic pillars and programmatic focus.

Strategy: Reach - Increase engagement and retention by employing user-centered design[edit]

This program included moving all of product to a user-centered focus in their development lifecycle.  We had two specific initiatives to complement that effort.

For the new readers research, we discovered that the research itself was incredibly valuable for ourselves but also for the community. The effort we put into having a quality presentation that we could share was rewarded and is a best practice moving forward.

We also kicked off Audience Segmentation research as a collaboration with Communications.  This effort focused on Audience Segmentations and will continue with a focus on research to help identify and support Mobile Editing development.

Strategy: Reach - Increase our global reach by increasing readership[edit]

People all over the world have yet to discover Wikipedia. We want to find out why[edit]

We went to India, Nigeria, and Mexico to reach the young, mobile, and search-dependent. We found that less than 50% of surveyed participants in all three countries recognized the Wikipedia brand or were aware of Wikipedia. In India and Nigeria, over 75% of participants said they had never heard of Wikipedia.

The Android app now offers new ways to browse Wikipedia and find trending, recommended articles[edit]

The updated Wikipedia app for Android features a completely redesigned home screen. It now includes the Explore Feed to not only help you find the information you need now, but also to provide interesting, recommended Wikipedia content to dive into. The improvements led to a 20% increase in the median time spent per session comparing the quarter it was launched to the previous quarter.  That new number has held.  The feed drives 12% as many pageviews as search on mobile apps (by far, our biggest impact to date).

How Wikimedia helped mobile web readers save on data[edit]

Photos are a ubiquitous element of Wikipedia’s most popular and highest quality articles, and this change means that your page will only load images as you scroll down, rather than on opening a page.

The new Wikipedia app for iOS puts the joy of exploration in your hand[edit]

Our new Wikipedia app for iOS features major design and usability improvements, including the new Explore Feed. The work contributed in a 54% increase in day-7 retention on the app (from 10% -->15.4%) upon releasing 5.05.

Additional initiatives:

  • Continued to improve user retention by implementing "Universal Links" (aka deep links) on iOS, which provide convenient re-entry to the app from links and OS search, but do not advertise or promote the app.  Android has had deep linking for awhile now, and with the improvements, iOS retention is now more or less on par with Android's rate.
  • Launched hovercards beta feature on desktop web across multiple wikis, gauge improved reader satisfaction via survey.  Our testing shows >20% increase in pageviews + hovers (v. pageviews alone).

Strategy: Knowledge - Maintain and improve content creation and editing tools[edit]

Making articles in all languages more accessible for readers[edit]

Compact language links replaces the long list with a short list of languages and a button that opens a dialog, which can be quickly searched to find other languages.

All-new notifications page helps Wikimedians focus on what matters most[edit]

A redesigned notifications page and a raft of new features make it easier than ever to manage and monitor your messages.

Cross-wiki notifications unite a world of messages in one window[edit]

The Wikimedia Foundation’s Collaboration team has developed “cross-wiki notifications,” a tool that will simplify the editing experience for Wikipedia users by allowing them to get updates from all of the projects they contribute to.  The work resulted in a 50% increase in usage for en.wiki and received many positive responses from the community.

Additional Initiatives:

  • Users can replace an image with a gallery that automatically advances through a series of images without taking up more space.
  • Provide Visual Editor as the default editing experience on all Wikivoyage wikis.
  • Single edit tab integration for Visual Editor for all users on several Wikipedias.
  • Provide Visual Editor as the non-default editor for anonymous users on the English Wikipedia.
  • Increase consistency of UI components within living (actual code) style guide.  The team made improvements to accessibility with color adjustments and worked across Design teams to have a common visual language.

Strategy: Knowledge - Maintain and improve content discovery[edit]

Opening the door to a new look: improving Wikipedia.org[edit]

We’ve improved the discoverability of information within the main Wikipedia.org portal to make it more contemporary and easier to use.  We reduced the amount of folks finding a dead end and no content by 14%.  Search usage from this page is up to 10% year over year.  User engagement, a new metric this year, through the portal also has increased by 5% and dwell time was cut in half. We also added informative metadata to the type-ahead search box, browser language detection, immediate display of translated page text with a re-sorting of primary links around the globe, along with numerous small backend changes to make the page display quicker and be more mobile friendly. Some of this work increased clickthrough rates on language links and sister project links as a result of visibility improvements. Overall page views to the portal increased from 15 million to 25 million.

Wikipedia seeks to speak your language[edit]

Wikipedia readers speak many languages, so it’s not a surprise that sometimes they search for phrases not in the language of the wiki that they’re currently reading. We’re changing that by adding language detection to Wikipedia’s search engine.

Supporting the future of Wikidata[edit]

The Wikimedia Foundation directly funds basic expenses for Wikidata software development. Wikidata is one of the youngest and largest members of the Wikimedia family. The free database has been edited by more than 16,000 users from all over the world. Today, Wikidata features 24 million items. Its data is being used 266 million times in its sister projects.

Introducing the unique devices dataset: a new way to estimate reach on Wikimedia projects[edit]

With the unique devices dataset, we’ve been able to quantify the shift to mobile across all projects. In almost all Wikimedia projects, more than half of our unique devices are accessing content using the mobile sites.  This effort was a collaboration between our Reading team and the Analytics Engineering team.

Additional initiatives:

  • Allow users to perform geospatial queries by adding support for geo-coordinate functionality into Wikidata Query Service.
  • Upgrading our search server clusters.
  • Wikivoyage sites switch away from Labs-based client code for maps and instead use Kartographer extension in the articles.
  • Improve relevance of intra-wiki search results, and a plan generated for switching to using cross-project indices.
  • Make the www.wikipedia.org portal a jumping off point to explore open content on Wikimedia sites.

Strategy: Community - Maintain and build community tools[edit]

Repairing broken links[edit]

Community and Wikimedia Foundation partnership with Internet Archive Library. English Wikipedia pilot to preserve copies of external pages used as sources on English Wikipedia, and when links are broken, provide links to archived versions. Over one million links have been repaired within three years.

Content translation tool hits milestone with one hundred thousand articles[edit]

Since its introduction in January 2015, the Wikimedia Foundation’s content translation tool has been used to translate 100,000 new Wikipedia articles. We asked some users to share their experiences about how this tool helped them.

How we partnered with volunteers to clean up copy-paste plagiarism on Wikipedia[edit]

In the four months more than 9,000 articles have been reviewed. Nearly 5,000 of them were found to be copyright violations and were fixed by patrollers. Thanks to all the new patrollers, there isn’t a growing backlog anymore, and new cases are reviewed within twenty-four hours.

Community Wishlist 2016[edit]

We recently compiled this wishlist. Teams will work with this list across Product, as well as the Community Tech team.  We also use the list as a focus at Hackathons throughout the year.

Community Wishlist 2015[edit]

Addressing the top 10 wishes on the wishlist, often in collaboration with other teams and volunteer developers, including the Wikimedia Deutschland Technical Wishes team.

Highlights include:

  • Wish #1: Migrate dead external links to archives - Community Tech supported a volunteer-run project on English Wikipedia, writing and testing code that helps InternetArchiveBot to detect dead links.
  • Wish #2: Improved diff comparisons -The nightmare diff in the proposal was a small change made in a huge paragraph, which highlighted the paragraph and obscured the change. A WMF developer updated the diff engine; now it shows what actually changed.
  • Wish #5: Numerical sorting in categories - Community Tech implemented support for numerical sorting and has deployed the feature to 18 different wikis so far; we're currently offering the feature to any wikis that want to try it.
  • Wish #7: Pageview stats tool - Community Tech built the Pageviews Analysis tool, which is now in use on all Wikipedias.
  • Wish #9: Improve the plagiarism detection bot - Community Tech created CopyPatrol, a new interface for the plagiarism detection workflow that's attracted more patrollers, and eliminated the backlog of cases.

More Team Updates[edit]

Fundraising Tech[edit]

The Fundraising Tech team did a magnificent job building on stability and performance.  That progress was on display this year. The teams made improvements to other areas and here are a few of those highlights:

  • Central Notice - Moved all new campaigns to use local storage instead of cookies and clean up years of cookie bloat. This is essentially "keeping up with the times," and it made the more technically savvy community members really happy.
  • CiviCRM - Deduping the database started and continuous improvements are ongoing. This is a huge win for tracking donor retention. MG is especially benefitting in saved manual labor and business intelligence.  Reporting performance and options greatly improved for all users. We are on a path to automate most of their regular reports. Civi Dashboard was fixed and enabled. This is heavily tied to reports and gives users quick access to information they care about.
  • Payments and processors - Successfully eliminated ActiveMQ (switched to Redis). The new system is a modern, scalable queue which handled the highest revenue day ever without any issues. This is the main reason we had 100% uptime this December. We expanded Astropay in 5 new LATAM countries and Adyen into Japan, Ukraine and Israel. It was also expanded as a backup processor for the US, UK and Canada. This is the first time we have ever had an English processor redundancy. Adyen also replaced Worldpay for France, a major win for stability and performance.

Team Practices Group[edit]

The Team Practices Group (TPG)  joined Product in 2016. TPG is a cornerstone for many cross functional programs.  The team facilitates and improve team off-sites, on-sites, scrum meetings, retrospectives, and planning. TPG  guided the Design team initiatives, WMF Values process, and Technical Collaboration Guidance discussions. TPG extended support for the Technology teams and processes, and reviewed and supported our Annual Plan process.

Program Collaboration[edit]

We saw growth in specific collaborative programs across our Product teams and with other departments. These programs include the Community Health Initiative , Audience Research and Structured Data on Commons. These programs will be a central part of this year’s Annual Plan.

In addition to regular roles and responsibilities, our Audience heads took on extra efforts.  Toby Negrin, Head of Reading, provided program support and leadership for Community Health & Anti-Harassment, Structured Data on Commons, Technical Collaboration Guidance, and Audience Research. Trevor Parscal, Head of Editing, supported the Check-in evolution and a fabulous All-Hands. Katie Horn, Head of Discovery and Fundraising Tech, represented staff for our ED transition and led the technology for our best fundraising campaign ever!

Communication[edit]

We also welcomed Victoria Coleman into the CTO role. We continue to work together with Technology and had a great syncing session in the fall facilitated by our TPG team.  Arthur Richards and Kristen Lans were excellent hosts and constructed one of the best cross department sessions.  We look forward to what they do with our Collab Jam planning later this month.

Well, that was a lot and of course we didn’t include everything.  We make improvements based on learning on the goals and objectives each quarter. You can always find out what our teams are up to by visiting the Product page on MediaWiki.  Please engage with the leaders of each team for more information and discussion.

We look forward to working together on the 2017-18 Annual Plan!