Wikimedia Foundation Annual Plan/Quarterly check-ins/Audiences1 Notes July 2017

Present (in the office): Katherine, Grace, Niharika, Abbey, Heather, James F., Danny, Toby, Trevor B., Zhou, Joe M., Neil, David B., Dayllan

participating remotely: Caroline, Ed, Erica, Joady, Maggie, Stas, Trevor P.


Slide 1 edit

https://meta.wikimedia.org/w/index.php?title=Wikimedia_Foundation_Annual_Plan/Quarterly_check-ins&wteswitched=1

Slide 2 edit

[No discussion.]

Slide 3 edit

time is a social construction - 3 wishes, #5, 6 and 7

Slide 4 edit

DH: Extra project we took on this Q, ACTRIAL-related. Helping enwiki concerned editors set things up. Hope to get some evidence so that we can move forward; opinions on both sides are very entrenched, we're trying to find out some reality, and build trust and co-operation. Cross-team collaboration from Community Tech, Research, Community Engagement, others.

Slide 5 edit

[No discussion.]

Slide 6 edit

DH: X tools are tools made by user X. For the people these tools are useful for, they're crucial. The author wasn't maintaining them any more.

Slide 7 edit

DH: Here's a more detailed, better version of this tool; thanks to Leon and Sam to make our users really excited.

Slide 8 edit

DH: Beta version live right now, making full transition soon.

Slide 9 edit

DH: Wikitext syntax highlighting helps editors.

Slide 10 edit

DH: I'm so happy to show this. Headers clearly marked, links and files in blue, italics/bold, references and templates highlighted. CodeMirror was created by a volunteer (User:Pastakhov), put in a lots of performance work which makes this something we can offer, unlike previous attempts.

Slide 11 edit

Even works in the 2017 wikitext editor. Lots of work by Niharika and Ryan Kaldari, and by Ed Sanders and James Forrester.

Slide 12 edit

DH: Live right now on Simple Wikipedia, going more widely soon (next week).

Slide 13 edit

DH: LoginNotify gives warnings in Notifications: one on by default (failed login attempts – if any from an unknown device or IP, or five+ from a known one), the other off but can be switched on (e-mail after successful loging from a new device or IP), aimed at sysops/functionaries.

Slide 14 edit

DH: Out next week.

Slide 15 edit

DH: This (grant metrics tool) was something we'd hope we'd get to last Q, but is taking longer. Part of the programs wishlist. Make things easier for grant recipients.

Slide 16 edit

DH: Top 10 in review. As well as getting Q4 stuff out, Brad J. is working on edit summary length (#2), and the team is working on #3 and #4.

DH: Wikitext syntax highlighting coming hopefully as a beta feature to all Latin script wikis next week, RTL and non-Latin scripts will take longer.

Slide 17 edit

[No discussion.]

Slide 18 edit

TB: Heya.

Slide 19 edit

TB: Welcome to new team members, now have people to build the things. David B and Dayllan M.

Slide 20 edit

TB: Went well. Some hold up on SuSA support.

Slide 21 edit

TB: Primary metric. Need to measure this. Want to avoid survey fatigue, working with SuSA to align with their work. Secondary metric will be qualitative, calling it "ANI" informally for now.

JM: Why the noticeboards?

TB: This is where complaints happen.

CS: Looking at overlap of admins. More of an official reporting pipeline.

KM: Very interested in the data.

Slide 22 edit

TB: Helping with admin tools. AbuseFilter complaints are lengthy, focussed on speed and richer controls for sysops to prevent harassment hopefully. Tool to understand the sequence of events should help see the pattern more simply and quickly, making the software do the work. Designs should be ready for development.

KM: Are there concerns around unintended consequences of making this more easily accessible? Making this more readily available, could that be problematic?

TB: Meeting with Legal around privacy concerns this week. We want to ensure that whatever else happens, we shouldn't provide tools that open up to more harassment. But should make things clearer.

TB: Page-specific blocking consultation to see if it's actually wanted.

TB: Mute feature allows private blacklist of people to not get Notifications and e-mails from; taking over from the Global Collaboration team.

EL: Looking at the screenshot, wondering if the nature of the privacy of the list should be highlighted in the preferences page?

TB: Yes, want to add link to the help documentation to make this clear. Lots of information about this to share with users.

Slide 23 edit

TB: Here's what the mute feature looks like

Slide 24 edit

[No discussion.]


Slide 25 edit

[No discussion.]


Slide 26 edit

TP: Working on reducing technical debt, unifying longer technlologies, increasing consistency between UIs. Part of that is taking risks. Looking into newer technologies like ML. Important to address specific things - making improvements to existing tools.

Slide 27 edit

JF: Focused on increasing UI consistency across tools. Worked on core library - make sure buttons etc are consistent across desktop and mobile.

Slide 28 edit

JF: Bigger program "visual" everything. Incremental improvements to curation tools, end-to-end consistent experience. Hoped to push out more widely, holding back for now. Getting great feedback.

VC: I'm not sure how you do the rollout?

JF: Yes, we can do beta features and get feeback from them. We can also do A/B tests with some things. For things like this, we want more interactive user testing. More individual user-research.

VC: We cleaned up the user research as part of org tuna. Can Abbey etc. help? Curious what flow looks like.

JF: We have Daisy, who does UR with Abbey and we work with Abbey too. Work with design colleagues across Audiences.


Slide 29 edit

JF: Improving experience with wikitext. Replacing Tidy to give a richer experience to users. But breaks a bunch of things. Now we have tools to help people identify things that will break things before they break so community folks can fix it ahead of time. We can use more modern tools with Remex. Based on HTML5.


Slide 30 edit

JF: Working through mobile content service project to improve quality of HTML we send out. Improved audio/video and redlinks support. Support for language variants coming soon.

Slide 31 edit

JF: Want to consolidate Content Translation tool with VE/NWE to make it available to all users. Will let us do Content Translation for Mobile. Continuing into current quarter.

EL: We're proposing using simpler language with users to explain things before they break.

Slide 32 edit

JF: Edit review improvements project. How people find them, how they figure out if they're good. Enabled by 30k accounts. Added more features. Preparing to be available in general soon.

Slide 33 edit

[No discussion.]


Slide 34 edit

TP: Doing a bit of research, wrapping up now. Abbey and Neil doing a lot of work, organizing everything. To better understand users we can't interact with directly. Research will inform details of things we get into here. We know device support is lacking. We aren't having to port a bunch of things, unifying first. Also continuing work around review and retention and keeping people engaged.

Slide 35 edit

TP: Reaching the singularity, unified editors, parsers, UI libraries across desktop and mobile. More efficient. Wrapping stuff up. Expanding support, bringing more tools to communities.

Slide 36 edit

JF: 2017 wikitext editor - aim to provide to all users. Rollout beta to more people. Will do user testing with mobile users. Have enabled non-JS editing on mobile. Great work by Jon Robson.

Toby: Can you talk about the rollout of beta to more users?

JF: Beta wasn't auto-enabled for everyone, not even if they check "Gimme all!". We have seen lots of edits made with it and are more confident with it now. Want to roll it out to more users.

Toby: One of the concerns I heard, are the gadgets and tools gonna ported/work with to new editor?

JF: No. We are working on helping gadget authors work on port them and talking to them if they even need them.

KM: What does this work look like?

TP: We went through a similar process for ... and we have an idea of what this looks like.

JF: Three prongs: 1. Documents. 2. What the gadgets are, what their impact is? 3. Testing and feedback. For example, HotCat is not WMF's but community doesn't edit it anymore so we need to have a conversation around if they still want/use it.

Toby: Time frames?

JF: Not really. Want to be confident first. Have done some user testing but need more before rollout.

KM: Are you the responsible for this process around community discussion?

JF: That would be Dan (Garry).

Slide 37 edit

JF: Much of it is reactive work. Have plans to remove 2006 wikitext editor. Have reached out to talk to some communities to get an idea of why they use it.

Slide 38 edit

JF: The UI library is not completely rolled out yet. Preferences page is the most complex one and it needs to work on mobile. Need to be careful about desktop-only features.

Slide 39 edit

JF: Hope to complete the Tidy to Remex migration this quarter.

Slide 40 edit

TP: The research will inform us to make targeted approaches which help us make significant gains.

Slide 41 edit

JF: Outcome of the research will be a report, conclusions. Hopefully coming by Tuesday. We can come up with possible solutions from the report, list of possible interventions. Want to discuss them before charging ahead.

Slide 42 edit

TP: Other half of dept's work is around editor retention. Important balance between demand and supply of review. Don't want to create more burden on users who are struggling to handle existing backlog. Starting with monitoring workflows, dashboard features etc.

TN: Have you learnt from ACTRIAL?

JF: I think, broadly, trust is not something you get magically, it's earnt. Say something, then do it.

TP: ACTRIAL is what happens when you don't maintain the balance. Sometimes community come up with solutions which might not be right. We want to address those pain points. ACTRIAL is a good case study for this.

Slide 43 edit

[No discussion.]

Slide 44 edit

JF: We want to provide the filters UI on watchlists too.


Slide 45 edit

JM: Significant re-design. We did it because with the amount of functionality we are adding; the old UI was already performing poorly with users, and wouldn't scale. Would become less usable. We needed a new paradigm. We will be extending to watchlists this quarter


Slide 46 edit

JF: Got feedback from survey, will plan to work on them. Dumps don't work with Flow so no research. Hope to triage and prioritise it and then work on it.

KM: Fuzzy on the arc of Flow.

JF: The scope was too broad and too many grand vision promises were made. Flow as in workflows might be something else in future, the current structured discussions tool will be more user-focused. I think we have a considerably successful product for some use cases.

JM: The survey gave us some clear priorities to make tool better for discussions. Work on the 5-6 main criticisms and concerns.


Slide 47 edit

Neil: Non-bots edit include bots because sometimes accounts don't have a bot flag.


Slide 51 edit

Neil: Active editors are generally flat.


Slide 52 edit

NQ: And these are our recurring metrics which we update every month on the Audiences page.

Slide 53 edit

TN: Welcome to Grace.


Slide 54 edit

AR: New Readers work continues, alongside the New Editors work.


Slide 55 edit

AR: New Editors work in Czechia and South Korea went well. Talked to a bunch of people. Lots of intersting experiences and perspectives. Details coming next week.


Slide 56 edit

AR: Worked with Reboot in the field, lots of literature review from English Wikipedia and wider studies, local researchers in the field helped translate not just language but also culture, non-vocal contributions. We provided the 'wiki' expertise.

AR: Now synthesising. Looking forward to sharing and broadening the findings. Working with Community Engagement and Programs as well as Audiences feels crucial, think it's key to achieving success.

KM: It's great, yes.

JM: Could also have an impact on ACTRIAL.

AR: Yes, could have some read across, we were looking not just at newbies but also at the users who are protecting the wiki.

KM: Want to highlight the issue around carrying capacity. Happy we're focussed on smaller and emerging communities, understanding how they've grown or not in relation to the size of the languages spoken.


Slide 57 edit

AR: This is the overall process. We're where the orange/yellow arrow is right now for New Editors.


Slide 58 edit

AR: Yay Gantt charts.


Slide 59 edit

TN: Welcome to Grace.


Slide 60 edit

GG: Working on improving communication, effectiveness, and agility across Audiences.


Slide 61 edit

GG: And this is the work done so far. Busy first five weeks.