A modest proposal for Wikimedia’s future
On February 25, Lila Tretikov, the embattled executive director of the Wikimedia Foundation (WMF), finally tendered her resignation. Though an interim successor would not be named until March 10 (see this week's Signpost news coverage), the Wikimedia movement breathed a collective sigh of relief.
Tretikov’s twenty-two month tenure produced the greatest organizational crisis in Wikimedia’s history. Her leadership will be remembered for poor communication, worse management, rapid and unannounced changes in strategy, and a lack of transparency that produced an atmosphere of mistrust and anxiety, one which finally overwhelmed and brought the Tretikov era to an acrimonious end.
Most of all, Lila Tretikov will be remembered for the precipitous decline in staff morale that sent more than two dozen key employees and executives for the exits. The loss of talent, relationships, and institutional memory is devastating, and it is not something the Wikimedia Foundation will recover from soon.
I suggest maybe the Wikimedia Foundation should not recover and rebuild itself, at least not exactly like it was. Acknowledging this modest proposal stands to be controversial (or, more realistically, ignored), I believe in this tragedy lies an opportunity for WMF to reconstitute itself in a way better suited for the challenges facing the Wikimedia Foundation at this point in its history.
This would be a WMF that recognizes its primary mission is educational, one that is willing to reconsider what responsibilities it keeps for itself vs. what works better distributed among its affiliates. I argue in this post that it should split its executive leadership into two roles and spin off certain core functions into standalone organizations. Doing so would allow for better transparency, create more opportunities for “WMF-Community” cooperation, and perhaps offer a chance for volunteers to seek a career path within the movement.
The Wikimedia Foundation does not need to do big things. It needs to create an environment for big things to happen.
The challenges to come
If the WMF is going to reconsider its organizational structure, this is certainly the time to do it. The forest fire of Tretikov’s tenure creates a unique and unexpected opportunity to plant anew. Other questions are already being explored: what will Wikimedia’s next five-year-plan say? Should Jimmy Wales continue to hold his semi-permanent seat on the Board? Are the processes for selecting and vetting the three groups of Board trustees still adequate, the underlying assumptions still operative? How can the Board be induced to act transparently? The Wikimedia Conference coming up in April should be interesting, if not explosive.
All of these are very difficult and important questions, and yet I strongly suggest opening another conversation about the size and scope of WMF responsibilities going forward. Why should the WMF consider radically re-envisioning its organizational structure? Because the WMF as it exists was created to solve a different problem than the one we have now.
When the WMF was launched in 2003, two years after Wikipedia’s creation, “Wikimedia” was a retconned neologism coined to describe a wide-ranging movement not yet fully baked. The WMF was needed to create a backbone for these efforts and give its global volunteer base a strong sense of direction. Under the direction of Sue Gardner, the WMF was successful in fulfilling this role.
The present WMF has become, in the pithiest description possible, a fundraising organization in support of a nonprofit web development company and a small-grant issuing organization. To a lesser degree, it has also funded community outreach and the development of membership chapters around the world.
Wikipedia, in its many languages and numerous sister projects—the larger Wikimedia movement with which this post is really concerned—has succeeded in becoming the world’s free resource for knowledge, however imperfect it can be. Maintaining this is a different kind of challenge, and it is inherently a defensive one. Indeed, there is much to defend, and the threats are not imagined.
The first challenge is the changing Internet: Wikipedia’s software and culture came from an Internet dominated by desktop computers accessing the World Wide Web. Today, Internet activity has moved to mobile devices, increasingly inside of apps, which are of course closed platforms. Though WMF’s mobile efforts have come a long way, they are fighting upstream against several currents no one imagined in 2001. The idea of collaboration is as strong as ever, but its tools become weaker all the time.
The second challenge is WMF culture. The Tretikov disaster reveals weaknesses in two of the WMF’s most important functions: the raising of money and the allocating of money. In addition, as described in varying degrees of detail by former staffers, under Tretikov the Foundation had become a toxic workplace environment—but the truth is it had structural issues even before that. Finally, the edifice of a nearly 300-person staff created a kind of intrigue—“Montgomerology” (a word coined by Liam Wyatt, referring to the WMF's address on New Montgomery Street) —that plays out daily on Wikimedia-l, a semi-public mailing list populated by Wikimedians, and lately the semi-private Wikipedia Weekly Facebook group, which this blog is frankly obsessed with. Which, I acknowledge, isn’t exactly healthy.
The third challenge, not unrelated, is Wikimedia culture. The English Wikipedia’s volunteer community, the movement’s largest and most influential bloc, is deeply set in its ways. Meanwhile, Wikipedia’s extraordinarily high profile contributes to a reluctance to tinker with, let alone radically rethink, how it conducts its business. And several bold initiatives developed within the WMF—including good ideas like the Visual Editor, debatable ideas like the Media Viewer, and bad ideas like the Knowledge Engine—have been received poorly by the community.
In all three cases, solving these problems are more than any one executive can handle alone.
The next steps
So what should happen? First, an apology from the Board of Trustees is definitely in order. Tretikov’s failure is entirely on them as Wikimedia’s ultimate corporate authority. Second, an audit / accounting of the failures of recent years. Wikimedia UK was required to do one following the Gibraltarpedia controversy; what’s good for the chapter is even better for the Foundation.
Third, the Board of Trustees should split the role of executive director into two positions: a president and provost, like universities do. Being an educational project, WMF should look to similar institutions for guidance. One becomes the “head of state”, handling the public and fundraising efforts, while the other handles administration and operations. Wikipedia’s high profile means that representing its value and values to the outside world is a full-time job. Regardless of whether Jimmy Wales remains a trustee, Wikipedia needs a new mascot, and it should identify a charismatic leader for this role, who may or may not come from the Wikimedia community. The provost position would be focused on grantmaking, community outreach, and long-term strategy. They must be a good manager and internal communicator, but need not be a big personality. And this person absolutely must come from the Wikimedia movement.
Fourth, and the really hard part, would be the voluntary dispossession of core Wikimedia movement functions from the central organization. The WMF should keep only what is mission critical—fundraising, grantmaking, legal, and communications—and spin off the rest. It has done this once before: that’s the origin story of the Wiki Education Foundation. WMF grants should fund these newly independent foundations, encouraging a reinvigorated support for community-driven organizations.
What is the basis for considering smaller organization sizes? From a theoretical perspective, there’s Dunbar’s number. The larger an organization becomes, the harder it is for everyone to know everyone else and understand what they’re doing. In the business world, this has been seen in the arrested development of agglomeration, once large corporations realized they had become slow and bureaucracy-laden. (Anyone else remember The Onion's “Just Six Corporations Remain”?) Critics of corporate consolidation were caught as flat-footed as the conglomerates they disdained when spin-offs became ever more popular. This is also an operating principle at Amazon, where they call it the “two-pizza rule”: “Never have a meeting where two pizzas couldn’t feed the entire group.”
From a practical perspective, the WMF’s behemoth status suits neither its day-to-day operations nor its perceptions by the wider community. As detailed by recently departed veteran staffer Oliver Keyes in The Signpost last month, systemic problems with hiring, promotions, and human resources in general were an issue at the WMF well before Tretikov’s arrival. Meanwhile, the WMF itself seems unapproachable, simply too much for anyone to wrap their heads around. Indeed the WMF itself is a conglomerate, of a kind. Creating more community space around its current departments would make each more accessible, generating more “WMF-Community” interactions. This would help greatly with transparency, and make it far easier to start new initiatives.
It all sounds pretty radical—and I’m not saying it isn’t!—but there are good reasons to think a new organizational structure could work. The argument against ultimately relies on an appeal to familiarity, bolstered by inertia.
The reorganization
With the caveat that I have never worked at the Wikimedia Foundation, nor in non-profit governance even for a minute, I won’t let that stop me from taking a crack at some specifics. What I write below is merely one way to go about it, and I encourage others—especially those with real WMF experience—to offer their view in the comments. Let’s go:
Among the WMF’s first major grants should be to the new Wikimedia Technology Foundation, containing the current Technology and Product teams. There is no critical reason why it needs to live in the same house as fundraising, and it would benefit from a strong leader with community ties—which it has not had for a long time. After all, even as we’re now sure Discovery is working not on a Google-killer but merely improved site search, it still ranks very low compared to other community-enumerated goals. Doing so will make its efforts more useful to everyday editors, and give it the latitude to develop for the next generation of Wikipedia editors. An early initiative of this spinoff should be to think about how to position Wikipedia for the mobile web and even consider partnerships with today’s media orgs—not so much the New York Times and CNN, but Facebook and Snapchat.
More complex would be the evolution of Community Engagement, encompassing grantmaking and outreach. WMF grantmaking has nearly always been hampered by thinking too small and funding projects too dispersed and under-staffed to be effective. Through its chapters, user groups, and various grantmaking committees it funds projects for not quite enough money which are basically nights-and-weekends projects, from which very few can draw compensation, thereby limiting their ambitions and achievements.
So while the core function of grantmaking should stay with under the provost at the slimmed down WMF, the bulk of its activity should happen outside the WMF. And the way this would happen is by the creation of a more ambitious grantmaking operation whose mission is to nurture and develop mini-foundations modeled on GLAM-Wiki US, the Wiki Education Foundation, and WikiProject Med Foundation. Rather than there being one new foundation, this needs to be a core capability of every mini-foundation that receives WMF funding.
Among the key projects necessary to healthy and functioning Wikimedia movement that could benefit from a devolved organization and dedicated funding: The Wikipedia Signpost, which is heroically staffed entirely by volunteers; the Wikimania conference, the locus of numerous organizational failures in recent years; Wikimedia chapter management: the model of volunteer support currently practiced focuses too much on geographic concerns at the expense of thematic topics, with considerable overlap.
Another might be content development: if you look at Wikipedia’s complete list of featured articles, it is arguable the only article categories supported by existing foundations are “art and architecture”, “education” and “health and medicine”, served, respectively, by the three model organizations listed above. Adapting from the list, this leaves dozens of top-level categories unserved by a formal organization, and decreasingly supported as the informal Wikiproject has withered in recent years. Very few Wikiprojects continue to thrive, and the ones that do—Military history and Video games—inadvertently perpetuate Wikipedia’s problems with systemic bias. By creating formal structures with specific outreach to associations and universities along these lines, Wikipedia can create more opportunities for outreach and collaboration.
What’s more, it would create opportunities for Wikimedians, particularly its younger cohort, to choose a career within the movement. Presently, there are too few jobs at libraries and museums to make use of all this talent. While conflict of interest (COI) issues will be justifiably considered, these fears are generally overblown. Nowhere in Wikipedia’s policies or guidelines—and certainly not in the Five Pillars—does it say that Wikipedia must be volunteer-only, and creating staff positions will actually reduce the likelihood editors will “sell out”. Wikimedia has long passed a point of diminishing returns on the volunteer-only model. And you know what? It isn’t entirely that now. We already live in a “mixed economy”, and we owe it to our community members to expand their opportunities. There’s no reason software programmers should be the only ones to earn a living working on Wikimedia projects.
Efficiency, transparency, and opportunity
Can I summarize all this in a paragraph? I think so: a small constellation of well-funded Wikimedia Foundation spinoffs, each with a strong sense of mission, focused narrowly on the movement’s needs stands a better chance of working more efficiently among themselves and offers many more touch points for the community itself to be involved. Through that, transparency can be improved, both at the WMF parent org and within a reinvigorated movement organized around professionally staffed, standalone foundations doing what each does best. In the gaps between them and the WMF, new opportunities for community involvement would arise for the benefit of all.
Wikimedia is vast, with an incredible diversity of talents and resources. It contains multitudes, and its organizational structure should reflect that.
This article was originally posted on the author's blog and is republished with his permission. The views expressed in this article are his alone and do not reflect any official opinions of this publication.
Discuss this story
She's an Aquarius ten years older than me! Wikipedia is young! --violetnese 23:08, 13 March 2016 (UTC)[reply]
Your own opinions, but not your own facts
Dear William,
so, I realize that the tongue-in-cheek headline indicates (at least to those who get the cultural reference or click through twice from the image) that these ideas and thoughts are not entirely serious. And to your credit you are candid ("won’t let that stop me") about a lack of expertise in the matters for which you are positing these solutions. Plus I happen to think that such a lack of expertise can be mitigated by keen observation and a good memory. Unfortunately though, the sheer amount of factual errors, distortions, omissions and misquotes in this piece makes me think that these mitigating factors are absent here, leaving these sweeping opinions and proposals on a very shaky basis. To highlight just a few examples:
With all that said, I agree that WMF should not be doing all possible work on its own. This has actually long been recognized, see the "Narrowing focus" strategy change implemented under the previous ED (a change which BTW has not been uncontroversial in the community, many felt that the Foundation should direct more attention to e.g. supporting GLAM work instead of just funding chapters to do so). The spinoff of the US and Canada education program in form of Wiki Ed is consistent with that, and they are doing excellent work. And the German chapter's success with Wikidata proved that some software development areas can be handled well by separate organizations. But cooperation between organizations comes with challenges of its own. Speaking of Dunbar's number (which BTW does not actually imply that successful organizations can't have 300 or more employees), it's worth being aware that successful cooperation between organizations in the movement has relied a lot on personal contacts too, facilitated by conferences, hackathons, site visits etc. E.g. I guess the Wikidata team would confirm that a good working relationship with WMF engineers - who operate the site their software runs on - has been quite important for Wikidata's success.
Disclaimer: My views here, while personal, are of course informed by my perspective as WMF employee, but also as a longtime Signpost author and former editor-in-chief who covered WMF activities from the outside (as main writer for "News and notes") for an extensive amount of time before joining the organization myself. Having written from a similar perspective, this make me appreciate that you are reading a lot of information, and processing it - which is actually not too easy a task. I have been reading your "The Wikipedian" blog at least since then, and linked to it myself repeatedly. That said, this is also not the first time that it contains glaring misconceptions that would have been evident to anyone familiar with the matter - such as when you informed your readers in 2013 that Jimmy Wales would leave the Board of Trustees by the end of the year (an error you graciously corrected after being notified, but I can't help seeing a similar pattern here of reading things into your sources that just aren't there. It makes me itch for the edit button to add lots of [failed verification] templates ;)
And speaking of the Signpost: I anticipate that the decision to syndicate this piece here will be defended with arguments like "but it was just meant to stimulate discussion" etc., and of course the Signpost should not shy away from publishing strong opinions and well-argued thought-provoking pieces. But even with op-eds, a certain minimum standard of factual correctness should be maintained. These topics are too important for the future of our movement to be covered at an accuracy level resembling that of a supermarket tabloid.
Regards, HaeB (talk) 00:37, 14 March 2016 (UTC)[reply]
PS, I could go on and on, but just one more example of hyperbolic catastrophizing combined with an apparent misinterpretation of sources:
Regards, HaeB (talk) 06:34, 14 March 2016 (UTC)[reply]
Spin-offs
The reasoning sound fine. As the WMF has more roles, it could be wise to have small organizations focused on a specific task.
However, this design can also have negative effects. It could be hard to know which activity corresponds to which organization. If power is too divided, the constellation could lack strength.
Also, if we have more organizations, who controls each? Currently the WMF is controlled by the Board of Trustees, a few of which are community-selected. With more organizations, would we have more Boards of Trustees? Would they report to the WMF? Would they be controlled directly of the Wikimedia community?
I'm not opposed to the proposal, I just want to prevent issues.
I agree. The technological requirements of the Wikimedia movement can be managed independently of the other tasks. Although, as we saw with the VisualEditor and Flow, the team must respond to the community.
That model is interesting, as the decision of which organizations to support is kept at the WMF, whereas any organization can apply to funds. It's not unlike the affiliates network.
I disagree: there is too little focus on "geographic concerns". Thematic organizations are great, but having only them would risk the worldwide coverage. Without Wikimedia Uruguay, nobody would care about the promotion of Wikimeida projects in my country.
For example, Wiki Ed covers Untied States and Canada only. Wiki Med does highlight the importance of multiple languages, and has members from multiple countries. But that seems more like an exception.
--NaBUru38 (talk) 20:33, 15 March 2016 (UTC)[reply]
Provost