Contributors/FY2016 Retrospectives/Department

Editing Department FY2015-16 (July 2015 to June 2016) high-level Retrospective. Attended by managers, team leads for all Editing Department teams. This retrospective meeting occurred after all individual teams had completed their retrospectives, and the top concerns from teams were represented.

Top Items and Follow-up ActionsEdit

  • Item: Were able to quickly respond to critical issues (production/engineering and community); we were able to get urgent/important things done, alongside the (Q) goals
    • Action to preserve: When doing Quarterly Planning, ensure goals remain intentionally limited so that teams have capacity to take on critical work — Trevor: "Commit no more than half your resources [to Goals]".
  • Item: Org stability from last few months is good and should be sustained.
    • Action to preserve: Prayer.
  • Item: The way we do Quarterly Reviews needs to change in response to the audience (ED and C-Level) changing and wanting something different (plus multiple specific concerns about Quarterly Reviews)
    • Action to improve: Get clarification from ED/C-level: what do they want to get out of Quarterly Review?
    • Action to improve: Wait for results from the Quarterly Goal/Quarterly Review Working Group (that Trevor is on) to fix this (but we are getting impatient)
  • Item: Lack of broader organisational/movement strategy makes it hard to work out what the direction for teams should be.
    • Action to improve: Trevor will involve people from department in strategy work in upcoming months, including at the departmental off-site in October.
  • Item: No new hires for some teams in over a year; workload not sustainable (especially for Language).
    • Action to improve: Teams should consider what support they would need in order to cut workload to match staffing
    • Action to improve: Runa to follow up directly with Trevor re: Language team workload and staffing.
    • Action to improve: At the Editing Department Off-site in October, the department will develop a prioritised skills/talents needs list so that any hiring opportunities that arise are used as well as possible.
  • Item: More/any advance travel and offsite planning [would be very helpful]
    • Decision: No teams should plan off-sites in June or July. (to avoid end-of-fiscal-year overlaps and conflict with Wikimania)
    • Action to improve: Trevor will put [early planning] on agenda for Editing Managers’ meeting next week.
    • Action to improve: Trevor to follow up with Wes on family support for travel.

Ground RulesEdit

  • Conversation this hour is confidential within the group
  • Written notes to be published to Editing and to general public after a review pass
  • Attendees represent both their teams and their own opinions

What are we [the Editing Department] trying to do?Edit

  • "We build collaborative, inclusive tools for creating and editing free knowledge."
    • Point: add “and maintain/improve”?
  • Make editing easier, more open and more pleasant (fun?)
  • Keep supporting our existing community of content contributors and curators
  • Accelerate the growth of smaller Wikipedia projects (and others?)
  • Increase participation with the assumption that broader, more inclusive participation yields broader, more inclusive and higher quality content.
  • Improve the MediaWiki platform, making it more suitable for our changing needs
  • Offer new methods of editing and participating on Wikimedia projects.
  • Be the partner, not the enemy or the troll, to communities looking to improve their editing experiences and workflows through technological change.
  • “Making the world a better place”

Review key events and results over the time periodEdit

  • Product releases
    • We released the visual editor to more Wikipedia projects as opt-in or default
      • (To almost all languages!)
    • Multimedia released cross-site uploads, ran into issues (on-going) with collaboration/support side of things
    • Deployment of Content Translation as a beta feature in Wikipedia in all languages was completed in July 2015. Many more features were added since than. It was used for creating over 100,000 articles.
    • Released cross-wiki notifications and concluded major upgrades of Notifications
    • Language had on schedule releases during the year with a better structured release and sprint timelines (compared to before)
    • Multimedia prototyped several new tools for contribution, e.g. ImageTweaks, FileAnnotations
    • Parsoid: performance improvements, improved editing support for VE, API cleanup
  • Data releases
    • Multimedia came up with first (ever!) iteration of real metrics for upload participation over time
    • Language created a public page for statistics about Content Translation.
    • General editing data releases (data warehouse work)
  • People
    • Multimedia: +½LA +⅟ₙBrion
    • Parsing: +Tim, –Marc, →Kunal
    • Collaboration: –Danny, +Benoît, +Joe, ←Kunal
    • VE: +David L., +Thalia, +½LA
    • Language: Niklas returned from half- to full-time
    • UI: +Volker
  • Parsing: Change of focus from Parsoid-only to Parsing in general
  • We survived executive churn
  • Events
    • Developer summit and All-hands
    • Most of our teams did their first ever off-sites this FY; others had a second/third
      • Parsing team offsite was very useful and productive and important for the team
  • Process
    • We gained UI standardisation responsibility/staffing (+Volker)
    • Team gained new insight via Phlogiston
    • Editing design meetings (Pau, Prateek, Volker & Ed)
  • Budget:
    • Stayed within travel budget
    • We made excellent use of nearly all of our contract services budget
      • Amir: this doesn’t say much to me.  What does it mean?
      • Trevor: Contract services is non-staff, non-Safeguard.  Often 0–10 or 0–20 hours, students or part-time. We were given budget w/o enough money for existing commitments.  By moving things around, and bringing people on part-time, were able to get a lot of value for the department and include some really great people in our work

Identify things that went well and should be preservedEdit

  • Relationships between teams
    • [RB] Language team is a good customer/user of Editing Design (Pau), Research & Data (Leila), and Analytics Engineering (Nuria) teams and think the interactions should continue as a core part of the process
    • [JF] Editing Design meta-team meetings (Pau, Prateek, Volker & Ed)
    • Collab team had good collaboration with Design Research (Jonathan); several timely, short-term research projects helped us make better decisions.
    • [RK] [JM] Collab loves its CLs (Benoît, Nick) as does VE (Sherry, Erica)
  • [SSS] Successful Delivery
    • [MH] Language focussed on analytics so that major impacts of projects are timely measured
    • [JF] [JM] [MH] [LA] Ability to quickly respond to critical issues (production and community); we were able to get urgent/important things done, alongside the (Q) goals. –Multimedia, VE
    • VE releases were generally low drama and, in pulling back when there was consternation, we managed to stay on track with broader work (at a cost of specific roll-outs/availability).
    • Parsoid deploys were also smooth and glitch-free
    • Some teams wrote excellent blog posts about their work
  • Other:
    • [JF] [MH] [LA] [Parsing] Org stability from last few months is good and should be sustained
    • Collab team feels their internal discussions are effective and working well.
    • Language has a routine of replying to feedback from users every day (but see also in the next section)

Identify things that should be changedEdit

  • Quarterly Reviews
    • [TP] [SSS] [JF] [RK] [LA] [RB] The way we do Quarterly Reviews needs to change in response to the audience (ED and C-Level) changing and wanting something different
    • Collab team thought QR meeting felt compressed, different teams taking wildly different amounts of time
      • Rehearsals?
      • Make longer?
  • Strategic Planning
    • [JF] [SSS] [MH] [AA] [LA] Lack of broader organisational/movement strategy makes it hard to work out what the direction for Multimedia (and, we imagine, other teams) should be
    • [JM] [JM] [RB] [RK] Collab team is not always sure how our team planning fits in with broader goals. In particular, as a team that does not have one specific toolset to focus on (e.g., visual editor), we can end up moving from project to project. Need to feel like there is bigger puzzle that the pieces fit into -- a five-year plan
      • I feel having a puzzle helps, but knowing what size and shape the pieces are will also be important –Mark
      • Runa: I voted for this because Language has a similar problem with multiple products
    • [AA] [AA] [JM] Status of Flow is uncertain; this is slowly being fixed but it’s difficult to find the time to do the research and product thinking necessary.
  • Logistical Planning
    • As a side note, from Multimedia, we had trouble getting all of our people in a meeting for a retrospective, especially since this process was very fast and one person went on vacation. More friendly async processes would be good.
    • Phabricator is restrictive and still doesn’t always support us in terms of giving visibility on goals.
    • [SSS] [ES] [AA] [RK] More/any advance travel and offsite planning
      • [AA] Any organizational support for families? (Ex: Scott expressed constraints on how many weeks he can take off from family, and there are multiple week-long events in the year - allhands, wikimania / hackathon, team offsite, department offsite).
      • More visibility earlier on travel budget
      • E.g. Product Managers' off-site organised with 11 days' notice (despite needing international and larger travel).
  • Relationships between teams
    • [SSS] [JM] [LA] Improve cross-team collaboration and projects (to enable big-picture thinking and bigger projects than possible with a single team)
    • Scrum of Scrums doesn't work for relationships with teams that don't attend.
    • DR availability very limited.
  • People, Resources & Staffing
    • [AA] [RB] [RB] [LA] No new hires for Language in over a year. Team wants to get more clarity about resourcing and when we can go back to ‘normal hiring’
      • (Perhaps “normal” = “growth-mode”, as during 2010-2015)
      • The org as a whole is not in "growth-normal" hiring mode right now and possibly never will be again.
    • [MH] [LA] Available Product Manager support is low for Parsing, Multimedia, UI Standardisation (all teams without a dedicated one).
    • Multimedia team is also highly frontend-focused, which is better staff-wise recently, but could be improved further with more resources (i.e. more hiring budget)
      • 3 front-end (Javascript, CSS) engineers and (not in the relevant fiscal year) one new back-end (PHP)
      • However, see above re: not clear what the organisational goals are, so maybe this is moot?
    • Language would love more clarity about support for translatewiki.net, which is an independent site, but closely related to what the team is doing and essential to the MediaWiki development cycle
  • External / stakeholder communications
    • [SSS] [JF] [MH] Parsing, Multimedia, VisualEditor teams: We should do more/some blog posts (or other similarly external channels) about what we do
    • MM, VE, UI: Challenges in other people in the Foundation knowing what we're doing.
    • Language receives a lot of user feedback (mostly about Content Translation, and also about ULS and other things), and this sometimes requires more time for replies and support than we planned.
      • Does Language have CLs?  No, Runa+Amir are mostly covering; these days getting more support from CLs (Sherry, Erica)
    • Language requests more non-US timezone centric meetings and events
    • [RK] [RB] [RK] Evidently our teams have drastically varying levels of CL support and integration
  • Better support for 3rd party users of VE / Parsoid
    • ? Why?
      • Lots of complaints / issues with installations on talk pages + on IRC channel + mailing list + phab tasks -- so could be just some time spent documenting / ironing out some kinks.
        • Oh, sure, but "why is this something we work on?". AFAIAA it's not in our "what we're trying to do" list.
          • We end up spending the time answering all those questions, doing releases. So, better to be efficient about it.

Discuss top items, including ActionsEdit

  • [JF] [JM] [MH] [LA] Preserve: Ability to quickly respond to critical issues (production/engineering and community); we were able to get urgent/important things done, alongside the (Q) goals. –Multimedia, VE, Collab, Language, Parsing
    • Any team felt blocked from handling a critical issue?
      • No, but James: still have a long-term lack of growth in editors.
    • PROPOSED ACTION TO RETAIN:
      • Ensure goals remain intentionally loose so that teams have capacity to take on critical work — "Commit no more than half your resources".
        • Trevor: commit only half of your resources to QGoals so capacity is available.
  • [JF] [MH] [LA] Preserve: Parsing: Org stability from last few months is good and should be sustained.
    • PROPOSED ACTION TO RETAIN:
      • Beyond our control
      • Prayer :-P
  • [TP] [SSS] [JF] [RK] [LA] [RB] Change: The way we do Quarterly Reviews needs to change in response to the audience (ED and C-Level) changing and wanting something different
    • Joe: specifically, that the audience has changed and the meeting has not changed to match the new audience’s needs?
    • Trevor: … need to adapt …
    • Joe: we were rushed, everyone just read their slides, meeting had little value over just reading the slides. Should meeting be broken up?
      • Roan: we’ve done rehearsals in the past.  That should help with timing.
      • James: decided not to do rehearsal because the two times we did we went over anyway, 1 time we didn’t we didn’t go over
      • Trevor: we didn’t speak to the audience the way they wanted.  Not sure how, but we need to do it differently.
      • SSS: It is not clear what purpose the QReviews are serving, what value to the org?
        • Trevor: ED and C-levels ask us to do for their benefit.
      • Ed: we can be a lot stricter on stopping teams at allocated time limit.
        • Runa: when time limit was set for each department, took into account teams presenting.  But after 2-3 QRs, we introduced metrics, which is taking a lot of time and we should have requested for time extensions or individual team times capped.
          • James: we don’t get to set the overall time, the ED does.
        • LA: I think time isn’t enough; need to talk to ED but if you have 40 people working on 20 things in 3 months, 1 hour may not enough time for in-depth.
          • But even with that, we get ~5x the time per staff member as Technology does.
      • Trevor: we are giving information that wasn’t valued and not including what they want. - we should only present what is needed and valued, nothing more
      • LA: now is a good time (with newly-perm ED) to ask what they are looking for
      • JF: put cost tag on presentations based on staff time spent to highlight the cost : value determination for those that make it.
      • Trevor is on a working group for QGoals and QReviews.
        • Roan: is that going anywhere?
        • Trevor: we are user testing
      • Should we push the C-levels to cancel it?
        • Joe: I think it’s good (for team) to have a review quarterly
        • SSS: It is not useful for us as a team, but probably useful for ED / C-level. It would be more useful if this served to foster discussion between teams / depts about goals for following quarters.
        • MH: not quite, but - in current form, not very useful.
        • LA: no, but should communicate that existing meeting feels rushed (which can lead to feeling dismissive of some teams/talking points) and requirements are somewhat arbitrary
        • Could we push for the Editing/Reading/Discovery QRs to be merged into a single meeting? The other C-levels only get one.
          • This would make things worse either way, IMO –MH +LA
          • And make it either a 3-hour nightmare or even more compressed? No thank you. The fact that our C-Level owns half the org is the org chart’s fault, not ours
            • More compressed == only Lead PM shows up and chats for 20 minutes == less work for everyone else, more of a high-level check-in and KPI discussion like it is for other C-levels.
        • What else is hierarchy good for? Team presents to vertical lead, presents to department lead, presents to ED. Skip the whole team having to present quickly, and the ED having to listen to everyone in the org. –MH
    • PROPOSED ACTION TO REMEDY:
      • Get clarification from ED/C-level: what do they want and what is valuable?
      • Wait for results from the QGQR (that Trevor is on) working group to fix this (but we are getting impatient)
  • [JF] [SSS] [MH] [AA] [LA] Change: Lack of broader organisational/movement strategy makes it hard to work out what the direction for Multimedia (and, we imagine, other teams) should be.  [merged with] [JM] [JM] [RB] [RK] Collab team is not always sure how our team planning fits in with broader goals. In particular, as a team that does not have one specific toolset to focus on (e.g., visual editor), we can end up moving from project to project. Need to feel like there is bigger puzzle that the pieces fit into -- a five-year plan
    • Trevor: I’m getting pressure from Wes and ED to have coherent strategy.  Focus was not new direction, it was keeping things going.  We lost nobody, shipped a lot of stuff, so that was successful.  This year, focus on [...]
    • PROPOSED ACTION TO REMEDY:
      • Trevor will involve people from department in strategy work in upcoming months.
  • [AA] [RB] [RB] [LA] Change: No new hires for Language (or Parsing or Collab or UI or…) in over a year. Team wants to get more clarity about resourcing and when we can go back to ‘normal hiring’
    • Trevor: We don’t have any planned hiring for this year.
    • Runa: Our team is facing burnout (based on staffing vs. workload), CX, maintenance work.
    • JF: VE have to had to scale back our commitments, yes.
    • Trevor: see a lot of people overworked.
    • Joe: just getting used to reduced team size, will see how it goes.
    • LA: constraint around putting out fires, involving community.  … not a crisis situation but limiting what we can do
    • Trevor: Could have this conversation (what do we need) at the department off-site.
      • Do we need more Product Managers?  Or engineering?  Research?  Design?  Already decided: our next hire is a designer with focus on research.
    • PROPOSED ACTION TO REMEDY:
      • Trevor: decide what not to do to reduce our workload to match staffing
      • Develop prioritised needs list at off-site
  • [SSS] [ES] [AA] [RK] Change: More/any advance travel and offsite planning
    • Trevor: Please don’t all teams plan off-sites in June
      • LA: Please don't plan anything in june/july (beyond Wikimania)
    • SSS: six-months' advance planning, especially helpful for people with families
    • Any support the org can give for families, so people can travel with families or work around their commitments
      • Trevor: I would 100% agree.  Would be wonderful if foundation could support families.  Not sure how much exec support or which avenue.
    • PROPOSED ACTION TO REMEDY:
      • Trevor: put [early planning] on agenda for Editing Managers meeting next week
      • Trevor to follow up with Wes on family support for travel.

Is this document sharable?  Yes, within Foundation as-is.  Will share to public after review/redaction managed by Phab task.

Next StepsEdit

  • Based on discussed items, collate action items for Department
    • Done, at top of document.
  • Share with Editing Department
    • Share to people who were in retros
    • Not to mailing list - that goes to dozens of other people too.
  • Publish
    • Joel to create tracking task
  • Read other retros
    • Publish other retros
  • Flag other items for specific follow-up
    • Joel will send doc to people in this meeting to share with their teams to follow up to see if there anything that somebody really wanted addressed that didn’t get addressed satisfactorily or at all.