Our desktop interface has been changing over time, yet since the introduction of the Vector skin (the default site design), most of these changes have been led by volunteers and are only available in prototypes, user scripts, gadgets, and volunteer-led skins. We think it's time to take some of these ideas and bring them to the default experience. Over the next couple of years, the readers web team will be researching and building out improvements to the desktop experience based on research and existing tools. Our goals are to make Wikimedia wikis more welcoming to new readers and editors, and easier and quicker to use for all (both newcomers and veteran editors).
This project is in the very early stages of discussions. There is no concrete plan or development yet. We need your collaboration and feedback on the general focus areas of the problem, and your thoughts on the initial design ideas below.
- 1 How you can help
- 2 Updates
- 2.1 January 2020: Language Switching Users Tests
- 2.2 December 2019: Prototype testing
- 2.3 October 2019: Technical Research pt 2
- 2.4 September 2019: Wikimania research report
- 2.5 September 2019: Desktop usage and behavior data analysis
- 2.6 August 2019: Research and brainstorming at Wikimania
- 2.7 August 2019: Technical Research
- 3 Problem statements
- 4 Goals
- 5 Constraints
- 6 Timeline
- 7 Metrics
- 8 Research and design process
- 8.1 Phase 1. Investigation and research, figuring out where we can create value, finding focus (May 2019 – September 2019)
- 8.2 Phase 2. Developing focus areas, sketching and prototyping ideas, starting conversations (July 2019 – November 2019)
- 8.3 Phase 3. Continued user testing and design refinements (October 2019 – January 2020)
- 9 Phase 1: Design ideas
How you can help
Feedback: Please look through these details and subpages, and share your feedback on the talkpage about the ideas and mockups we have collected so far - How could they be further improved? What important considerations are not currently documented?
Research and ideas: Please edit the "Phase 1: Design ideas" sections below, to add new ideas, notes, links to past discussions, links to existing gadgets/scripts, links to good external design-patterns to consider, etc.
Individual testing: Once development begins, we will make the ongoing work available directly via a new skin option in Preferences. You will be able to individually opt-in and test the changes.
Whole wikis: We are also looking for any wikis that are interested in helping with widespread testing, such as changing the local default for logged-out site users. If you think your community might be interested, please ask us if you have any questions, then start a local discussion, and once there is local consensus add a link to that discussion on the talk page.
Longer comments and thoughts are welcome on the talkpage, in any language.
Subscribe to Newsletter:Desktop Improvements updates to be notified when new sub-sections are added here. Major milestone announcements will go to all wikis via MassMessage.
January 2020: Language Switching Users Tests
In December we did some usability testing to compare the current placement of the Universal Language Selector (in the sidebar) with the proposed location in the article header. We tested a prototype with 21 users in order to determine if people have an easier time switching languages given the new location of the language switcher. The results of the test confirmed our hypothesis — participants in the test group (new location) were able to switch languages more quickly than participants in the control group (old location). Based on these results we plan on continuing to explore moving the language selector to this location as a part of the desktop improvements project.
December 2019: Prototype testing
Between December 2019 and February 2020, we will be performing tests on a prototype for the first few features of the desktop improvements project. We will be gathering feedback from a variety of test wikis by encouraging editors to participate using a central notice banner. We ran the banners and received feedback from the majority of our test wikis during December 2019. In February 2020, we will continue running the banners on English and Polish Wikipedias. So far, the feedback we have received is mostly positive but we have also identified some areas of our prototype that we will iterate on based on the feedback. We will be publishing the results of the first round of feedback over the next few weeks, and of the second, sometime in February. In the meantime, we encourage you to give us feedback (if you haven't yet) on the prototype page.
October 2019: Technical Research pt 2
We have published the results of our technical research in preparation for beginning the work on the project. In particular, we have focused on gathering more information on the following:
- The possibility of serving a different version of the site to logged-out users (for opt-in/opt-out and A/B testing purposes)
- Search and the differences between the current search widgets
- Language switching and possibilities on improving the language switcher
- T234907 - A technical RfC on what code-base to start with
- Changes to EventLogging and how they will impact our work
September 2019: Wikimania research report
During Wikimania 2019, we interviewed editors with the goal of sharing the plans for our upcoming Desktop improvements project, and collecting valuable feedback on a number of preliminary design ideas. Our research consisted of user interviews, a free-form feedback exercise, and a presentation with breakout groups for more focused discussion. We have published a report and a PDF summary (in English) of the feedback we received. Overall, we received positive feedback on the focus areas selected, as well as the individual prototypes for ideas. However, we were also able to identify areas for improvements.
We will be iterating over this feedback over the next few weeks and plan on developing a prototype that we can test with a wider audience across wikis.
In the slideshow below is a sample of 19 of the ideas we tested. For more context, please read the full report and then give us feedback on the talk page!
September 2019: Desktop usage and behavior data analysis
As a part of our research process, we wanted to learn more about the way people currently use the site. In particular, how often they use available functionality such as links in the sidebar, language switching, and search. We have published our results. Overall, usage of sidebar links is low - only about 0.5% of all logged out users and 1.6% of all logged-in desktop users clicked on one of the pages linked in the sidebar. Language switching usage varied, generally based on the size of the wiki, with smaller wikis switching languages more often.
August 2019: Research and brainstorming at Wikimania
Wikimania provided us with the opportunity to speak with experienced members of our communities. Over the five days of the conference we were able to share the plans for the project, and collect valuable feedback on a number of design ideas. Our research consisted of user interviews, a free-form feedback exercise, and a presentation + brainstorming session (you can see the slide-deck we used at the side). We will soon be publishing the summary of the feedback, as well as interview session results.
August 2019: Technical Research
To explore different technical possibilities for the project, the team spent a week hacking on different approaches to a single problem - how to enable the sidebar in the desktop experience to be collapsible. We are hoping to use the results from these experiments to determine the technical architecture for the improvements, as well as the skin we would like to build these improvements within. Here is a list of the experiments themselves:
- Building a collapsible sidebar using a fork of the vector skin
- Building the desktop sidebar within the Minerva skin. An exploration in the question: "is there a world where we have the same skin for desktop and mobile?"
- Building a collapsible sidebar using user styles only
- Wikimedia wikis do not feel welcoming
- Wikimedia wikis’s desktop sites are not a welcoming or familiar experience for new readers. It does not match the expectations created by the modern web and our other platforms (the Android and iOS apps as well as the mobile website). It feels disorienting and disconnected due to the haphazardly organized navigation and interface links. In turn, this causes readers and editors to have less trust in Wikimedia wikis, to be less likely to explore Wikimedia wikis, and eventually, to use our sites less than they otherwise would (i.e. a decrease in retention).
- Wikimedia wikis are not easy to use
- Readers (new readers especially) are unable to intuitively perform basic functions like switching languages, searching for content, or adjusting reading settings. Additionally there is a lot of clutter that distracts from the content they are interested in. New editors are faced with a similar barrier; an interface that is not welcoming or intuitive and is cluttered. It is difficult for them to perform basic tasks necessary for contribution, such as setting up an account, opening the editor, or learning how to use special pages for moderation purposes (e.g. history pages to find&revert vandalism). All users can have problems with the sites not being properly "responsive", which leads to issues like really wide content or really narrow content (depending on screen and window sizes). By keeping the status quo, we are preventing people who are eager to contribute from being able to do so (i.e. experience bias).
- Wikimedia wikis’s model is not easy to understand
- Currently, a very small percentage of readers understand how Wikimedia wikis function. Our interfaces do not highlight the inner workings of the site in an intuitive way. Many readers are not aware that the content they are reading is written by volunteers and updated frequently, or that they can potentially contribute as well.
In addition, the large difference in experiences among our various products (desktop, apps, and the mobile web), makes it difficult for readers to know the connections between our products and to associate them with the content itself. This creates a lack of unity in the concept of Wikimedia sites.
Summary by audience:
|New readers||Not a welcoming experience — overwhelming, disorienting, unfamiliar, jarring, not trustworthy|
|Existing readers||Not an ideal reading/learning environment — cluttered, lack of emphasis on content, lack of emphasis on popular actions, opaque, lack of emphasis on depth & breadth of knowledge|
|New editors||Difficult to get started — intimidating, confusing, disorganized, exclusive|
|Existing editors||Not an ideal editing/community environment - lack of emphasis on useful actions, disorganized tools, many tools given prominence are not helpful|
Here are the outcomes we're working towards:
- Make it easier for readers to focus on the content
- Provide easier access to everyday actions (e.g. search, language switching, editing)
- Put things in logical and useful places
- Increase consistency in the interface with other platforms - mobile web and the apps
- Eliminate clutter
- Plan for future growth
Here is a list of things we would explicitly like to keep in mind:
- Not touching the content - no work will be done in terms of styling templates or to the structure of page contents themselves
- Not removing any functionality - things might move around, but all navigational items and other functionality currently available by default will remain
- No drastic changes to the layout - we're taking an evolutionary approach to the changes and want the site to continue feeling familiar to readers and editors
Here is the roughly planned timeline, which will change based on progress:
- Phase 1: May – September 2019: Investigation and research, figuring out where we can create value, finding focus
- Phase 2: July – November 2019: Developing focus areas, sketching and prototyping ideas, starting conversations
- Phase 3: October 2019 – January 2020: Continued user testing and design refinements
- Phase 4+: To be determined
Below is a draft of the core metrics we want to measure through the course of the project. As we define the changes we want to make with more specificity, we will expand and iterate on this list.
Increase utility among our existing audiences, proxied by:
- Increase searches per session by 5% over the course of the project
- Increase language switching per project by 5% over the course of the project
- Increase in positive and welcoming sentiments towards the site (via surveys and user testing)
- Increase in sentiments of trust and credibility (measured via surveys and user testing)
Research and design process
General note: our process is not particularly strict. It is based loosely on research and design process best practices, however it is also relatively emergent and flexible, in that we are engaging in ad hoc activities and explorations as they feel appropriate. While we've outlined the research and design process in three phases below, in practice the phases overlap. Additionally, for the time being there doesn't seem to be a clear way of separating research activities from design activities (they are interrelated) so we're discussing them together.
Phase 1. Investigation and research, figuring out where we can create value, finding focus (May 2019 – September 2019)
- Main page: /Research and design: Phase 1
We began by considering the current default experience on desktop (Vector) and asking ourselves: in what ways can we improve upon this? Where are opportunities to modify the interface in order to create a better experience for all readers and editors? How can we make it easier for people to do the things they want to do? How can we create a more pleasing reading environment? Of course while exploring these questions we kept in mind the project's constraints. The research and design activities we engaged in in order to explore these questions included:
- Understanding the history of the desktop interface
- Reading previous Wikipedia research conducted by the Wikimedia Foundation or other research institutions and individuals
- Discussions at our team offsite to develop a shared understanding of the project and generate ideas
- Winter, Timeless, and other Wikipedia redesigns
- Reading about redesigns/updates of other large websites (Reddit, Twitter, etc.)
- Conducting an audit of other large websites to try and glean common structural elements
The output from phase 1 was: a better understanding of the desktop interface, and proposed focus areas for improvements. We didn't have strict criteria for what a focus area could be. Generally speaking it was an idea of an improvement we could make, though at varying levels of specificity, such as: a less cluttered reading experience, or language switching ease. The proposed focus areas are:
- Creating a more focused and "quiet" reading environment by consolidating or optionally collapsing navigational links, including:
- Main sidebar navigation
- Article tools
- User tools
- Language switching
- Article navigation / table of contents
Additional, more feature-specific, ideas that came up: reading preferences (e.g. dark mode), share button, larger edit button / add new article button (for smaller wikis) / making it more obvious how to "get involved", article stats / activity summary.
Phase 2. Developing focus areas, sketching and prototyping ideas, starting conversations (July 2019 – November 2019)
- Main page: /Research and design: Phase 2
The research activities and conversations in phase 1 helped us develop a better understanding of the landscape we were working within (i.e. the desktop interface). It also helped us develop potential focus areas to further investigate (while still remaining open to new ideas). Our next goal was to dig a bit deeper into the focus areas through sketching, prototype, and most importantly conversations with the community. The research and design activities included:
- Understanding past work, research, and experiments in respective focus areas
- Obtaining general usage data about the respective focus areas
- Sketching out and prototyping early ideas to help facilitate conversations
- Forming early hypotheses
- User interviews and other feedback at Wikimania
- Community feedback via MediaWiki (happening soon)
- User interviews with newcomers and casual readers (happening soon)
- User testing on usertesting.com (happening soon)
The output from phase 2 will include: reactions to sketches of specific interface improvements within the focus areas, a refined (i.e. more opinionated and informed) sense of which improvements are worth pursuing, a proposal for the sequence in which we could implement proposed improvements.
Phase 3. Continued user testing and design refinements (October 2019 – January 2020)
Phase 3 will be a cycle of: further testing the specific ideas that came out of phase 2, then refining our designs in response to what we learn. There are potentially some things that need to be tested as betas on actual wikis. We will be working to identify these things and figure out how we plan to test them in beta (i.e. what data we would be tracking, and what decisions we'd be making based on what behavior we see).
The output of phase 3 will include: nearly finalized designs (though we usually leave room for additional iterations during implementation) and a plan for what we want to learn as we roll things out in beta, and what decisions/changes we would make based on that information.
Phase 1: Design ideas
A few of the many ideas that have been suggested are below. Please add any prominent ideas and links you know of:
- NOTE for editors: Please edit the English version to add ideas and links, or comment on the talkpage. Don't worry about the "tvar" markup if you don't know it. We will add it in afterwards.
Focusing on the content, distinguish content from user-interface
- Collapsible sections in sidebar
- Collapsible sidebar
- Floating sidebar
- Examples: FloatSide (enwiki script), ...
Easier access to everyday actions
- Sticky header with search, table of contents, edit links
- More prominent language switching (moving the language switcher to the top of the page)
- Examples: Timeless skin at 1325px or wider and at 1085px or thinner, ...
- Sticky table-headers
- Examples: Gadget-StickyTableHeaders (enwiki gadget)
- Jump to the top of the page
- Examples: ToTopButton (enwiki script)
Putting things in logical and useful places
- Consolidated user menu (e.g. collapsing things like "Log out", "Preferences", and "Beta", inside a dropdown menu)
- Examples: Compact personal bar
- Preferences for logged out users
- Examples: Accessibility settings/preferences (T91201), ...
- Moving article actions from the sidebar to within the article
- Examples: Winter, Timeless skin at 1085px or thinner, ...
- Easily navigate through articles - load page content using JS in background without loading the whole page