Next week is Wikipedia’s 15th birthday, the first draft of the long awaited strategic plan of the Wikimedia Foundation will be published for comment, and yesterday was the start of its annual “all staff” meeting. Meanwhile… there is a battle going on at the top for its soul.
- staff morale at the WMF is at an all time low (only 7% feel informed, only 10% feel confident in senior leadership, as reported in The Signpost);
- over the last year, there has been extreme “churn” in senior staff (as noted by William Beutler and MZ McBride and most notably by former WMF Director of Features Engineering Terry Chay);
- there has been a “transparency gap” including the mid-2015 major, yet secretly planned, “re-org” of the engineering department (see the list on the talkpage of WMF.WTF);
- last week was the unprecedented dismissal of James Heilman, a community-recommended member of the Board of Trustees who I personally know and trust, due to a loss of “mutual confidence”. The decision was persuasively defended by Denny Vrandečić, a community-recommended member of the Board who I also know and trust, but with insufficient justification for many – including the longest-serving employee of the WMF, Tim Starling;
- this week, in a blow in terms of diversity-of-perspectives, two American, San Francisco bay-resident, tech industry insiders, have been appointed to the Board – one of whom (appointed specifically for his HR experience) was named in a US Department of Justice investigation into anticompetitive wage-fixing cartel among tech companies, as comprehensively summarised by Jim Heaphy.
WMF’s Executive Director Lila Tretikov with Board member Jimmy Wales
It is my supposition that this is not a list of unrelated incidents, but that this is part of a wider theme: That a portion of the Board of Trustees and the Executive Director of the Wikimedia Foundation believe that it should be treated as a technology organisation in the style of a dot-com company, out of step with the staff and without the awareness of the community. By contrast, it’s always been my belief that the Wikimedia Foundation is an education charity that happens to exist primarily in a technology field. Of course software engineering is crucial to the work we do and should represent the major proportion of staff and budget, but that is the means, not the end.
All this background makes next week’s WMF draft Strategic Plan a very important document. For the 2010-15 plan there was a massive community consultation project but this time around there was only a 2-question survey. As Philippe Beaudette, the Community Facilitator on that original strategy process and latterly the WMF Director of Community Advocacy (who also recently left the organisation), said to me [with permission to publish here]:
The Wikimedia Foundation has one unique strategic asset: the editing community. Other orgs have great tech resources, tons of money, good software, and smart staff… but none of them have the editing community. I am, frankly, saddened by the fact that this one unique strategic asset is not more central to the developing strategy.
The November staff presentation gives a strategy preview that speaks of three priorities (slides 28-30): “1. Engage more people globally (reach) 2. Facilitate communities at-scale (community) 3. Include broader content (knowledge)”; and describes a need to “prioritise core work” (slides 32-33). All laudable goals, but they only include “example objectives” such as “build capacity”, “improve trust”, and “improve tools”.
Nevertheless, I suspect that the major strategic direction has already been privately determined. In short, it appears there will be an attempt to create the internet’s Next Big Thing™ at the expense of improving the great thing that we already have.
- In May, as noted by Risker, “Search and Discovery, a new team, seems to be extraordinarily well-staffed with a disproportionate number of engineers at the same time as other areas seem to be wanting for them.”
- The June staff presentation “strategy preview” talks about creating a “knowledge engine where users, institutions and computers around the world contribute and discover knowledge”. The FAQ page for the “Discovery department”, describes this project as “…improving the existing CirrusSearch infrastructure with better relevance, multi language, multi projects search and incorporating new data sources for our projects.”
- In September the Knight Foundation gave a grant of $250,000 to build a “knowledge engine”. This was announced by the WMF two days ago. This is a “restricted grant” but, as has been described by Pete Forsyth, there is none of the associated documentation – for example the formal grant deliverables – except for a short FAQ.
- As mentioned above, we now have two new Silicon Valley executives appointed to the Board of Trustees. They join the previously appointed member of the board Silicon Valley venture-capitalist Guy Kawasaki, as well as internet entrepreneur Jimmy Wales himself. There is no one appointed for their professional experience in education, charities, communities or developing countries.
While I agree with the general premise that the search system on the Wikimedia projects can be improved, I don’t know anyone who thinks “an indexed & structured cache [of] Federated Open Data Sources” should be THE strategic priority. Starting something entirely new like Federated Search is HARD and trying to include external sources (that link also suggests trying to also integrate the US Census, and the DPLA) is even harder, especially when there are so many existing technical needs. Quoting Philippe again; “for instance, fixing the inter-relationships between languages and projects, or creating a new admin toolset for mobile, or paying down our technical debt, or establishing a care/command/control operation for critical tools to ensure their sustainability, etc….”.
The Funds Dissemination Committee (on which I sit as a community-elected member) declared in November that it is “…appalled by the closed way that the WMF has undertaken both strategic and annual planning, and the WMF’s approach to budget transparency (or lack thereof).” In response the WMF is considering submitting its 2016-17 annual plan, based on the aforementioned strategic plan, to a “process on-par with the standards of transparency and planning detail required of affiliates going through the Annual Plan Grant (APG) process”.
We will see over the next weeks to what degree the apparent shift towards a Silicon Valley mindset – whether the staff and community like it or not – is indeed true. As the then-Chair of the Board Jan-Bart de Vreede said in describing Lila Tretikov’s appointment as Executive Director,
We are unique in many ways, but not unique enough to ignore basic trends and global developments in how people use the internet and seek knowledge…I hope that all of you will be a part of this next step in our evolution. But I understand that if you decide to take a wiki-break, that might be the way things have to be.
Meanwhile, you might be interested in this three-year roadmap for the Discovery department, for the more technically minded there is the “Discovery” workboard on Phabricator and associated mailing-list. Finally, for what it’s worth, the term “knowledge engine” itself is now deprecated.
[Edit: Since publication, this blogpost has been linked from
- The Washington Post, Wikipedia just turned 15 years old. Will it survive 15 more? by Andrew Lih.
- The Wikipedian.net, The Crisis at New Montgomery Street by William Beutler
- Heise Online, 15 Jahre Wikipedia: Die Unvollendete by Torsten Kleinz]
Witty's Blog 
You hit the nail on the head, Liam. Thanks for writing this.
Thanks Liam for writing this blogpost!
As for the “churn”, I don’t see why it would necessarily be bad. A lot of the people who left WMF were seriously harming the projects. Community members complained loud and clear about multiple software engineering projects, it was only appropriate for some of the responsible individuals to go away. Especially those which treated the Wikimedia wikis and community as their personal playground.
I agree with your premises, but I’m skeptical of your conclusion. Two additional points to consider:
(1) Where staff have most visibly harmed the projects, they have done so with the support of the ED and the board. I believe Superprotect is the clearest example; 1,000 people signed the letter I wrote to the ED and the board, but the recipients have declined to even acknowledge the letter’s existence, nearly a year and a half later.
(2) The staff survey was conducted in November 2015 — just a couple months ago. If your theory is that she has been ridding the WMF of “bad eggs,” she’s had a year and a half to do so. The people who had left WMF in her first 18 months (whether willingly or unwillingly) were not among the ~90% who apparently have concerns about senior leadership.
(Addendum to #2: It would appear that the many staff that have been *hired* under her leadership *are* among that ~90%, though — considering that the staff size has been both churning and growing during her tenure.)
(1) Not really true. For instance Flow was halted by Lila, VisualEditor was slowed down by her, Gather was put on hold. There are more examples.
(2) Most of the people who harmed the Wikimedia projects are still there (a good portion of WMF staff, if not a majority, worked on things which harmed the projects). If you count the people who worked on some of the projects which Lila put on hold, you’ll see they’re still dozens and they probably worry about the future of their own seat (because they should).
Churn is bad…always. Either it means you made a bad hiring decision or its a loss. At the minimum the time lost hiring a replacement and bringing that person up to speed is exorbitantly costly. In the tech world, headhunters can take 3-4 months an engineering salary for a single placement.
…
Three years before I became a freshman in college, the house (Caltech’s version of a dorm-meets-coed-frat) I ended up in decided that they would “lock in” all the “cool” freshmen during rotation (a mandatory rush week) by being assholes to the “non-cool” ones. This idea backfired because a lot of good freshmen get swept up in this—they understandably don’t like to see their “uncool” friends being treated like shit and opted-out of the house.
If there is one thing my analysis of engineering and product employee retention at the WMF shows is that common-sense concept writ in a commercial setting. Removing someone over one issue will have repercussions in the relationships and morale of all the others that person touches, even if their involvement is ancillary to the issue at hand. This ripples out into a cascade of unintended consequences. The more entrenched or important that person, the larger the cascade.
In other words, do you think that WMF staff members left or were fired because community members complained, or because of differences with people they work and interact with on a daily basis?(*) How much input do you think a typical staff member (even a director) has on the sorts of strategic decisions that might piss off a community member—and yet it’s those staff members that seem to be quitting or expressing dissatisfaction.
Finally, which staff members treat Wikimedia wikis and community as “their personal playground.” I knew of only one incident where a staff member deliberately abused admin rights on a wiki (english Wikipedia), and that staff member was eventually fired over it.
…
In any case, I’m glad I left long before this stuff happened. 😀
(*) AFAIK the only staff member I am aware of who was fired through the community (and not by the staff) was the single staffing decision that touched off all this churn in the first place. That’s because the only direct action the community can take on the WMF is through the actions of the (mostly community-elected-but-I-guess-thats-changing) Board. Board decisions only affect the executives (obviously), can only happen once or twice a year at most, and can often take over a year before it rises to that tipping point. The old trope: “Be careful what you wish for, community, you just might get it!” (And you did.)
> Either it means you made a bad hiring decision or its a loss.
As far as I can see, most people who went away during Lila’s tenure were hired before she arrived. And by the way, Terry, you are one of those who are not missed.
Thanks a lot.
Liam, so much to address here — you make many good points. One thing comes to mind, regarding your first underlined statement above (Wikimedia Foundation as technology organization). In her first major interview after being hired (which, for what it’s worth, badly mis-characterized my own views, though the reporters showed no interest in communicating with me before or after publication) — Lila Tretikov’s vision for Wikipedia was characterized as “to assert itself as a technology company, rather than a media firm.” This statement has always struck me as a very strange dichotomy. I don’t think anybody on the planet would say WMF should be a “media firm,” but that’s a far cry from asserting that it should be a “technology company.” Today, though, those words seem very telling; the pursuit of the Knowledge Engine, planning it behind closed doors, bringing in new Silicon Valley-based Trustees… it’s all pretty consistent with her background in technology, and her interest in artificial intellingence. http://www.theguardian.com/technology/2014/aug/06/wikipedia-lila-tretikov-glasnost-freedom-of-information
Thank you for writing this, Liam.
Thank you for all the information and links, it draws a big picture not always clear for people outside. Some reflection about it is needed.
Pingback: Meltdown at Wikipedia? | Posthegemony
Thank you for blowing a whistle, never mind how feeble or harsh it may sound like.
Liam, thanks for this recap. Two comments:
“Of course software engineering… should represent the major proportion of staff and budget”
— Why ‘of course’?
“I suspect that the major strategic direction has already been privately determined.”
— From my experience, this is not so. To the contrary, neither the WMF nor the movement seem to have a strong strategic direction at present, nor even a set of clear options to choose from. That’s why discussions about what it /might/ be are so vague.
Sure. On the other hand, did the WMF official strategy ever really matter to WMF? WMF never achieved anything about its own strategy. in 2013, WMF stated: “We have known for several years that we will not achieve the 2015 plan targets” https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Annual_Plan/2013-14#EXECUTIVE_SUMMARY
Exercise for the reader: imagine how this piece of news would have been given under another ED: https://blog.wikimedia.org/2015/09/25/wikipedia-editor-numbers/ (I still have no idea!).
Pingback: » The Crisis at New Montgomery Street The Wikipedian