I'd like to add some thoughts to the discussion about the potential pros
and cons of spinning parts of the Wikimedia Foundation off. I’m writing
this in my personal capacity and this email might not represent the views
of the Wiki Education Foundation.
This is a comparably long note; the upshot is that in my opinion there are
more pros than cons.
== PRO ==
* Distributed risk: If one part of the new ecosystem fails, the rest should
still be healthy enough to survive. Today, if the WMF implodes, everybody
else will be affected in a big way. If we manage to successfully create a
number of separate organizations of what is today WMF, the risk will be
spread. Some parts will still be vital for the survival of others, though.
So, that has to be taken into account during the process of spinning off.
* Specialization of EDs: People who are experts in their field tend to
produce better results. During the last ED search, WMF struggled with
finding a “unicorn”. In order to run the WMF, you’ll have to be able to
deal with a high level of complexity, understand the community, be willing
to deal with public criticism, understand how to build an effective
engineering organization, etc. Splitting WMF into different organizations
would make subsequent ED searches easier as their expertise won’t have to
be as broad as under the current conditions.
* Better focus: Organizations with a narrower focus are likely to do
better. Wiki Ed is only one example. I’ll leave it with that because LiAnna
has explained this point already nicely.
* Shorter and more efficient decision making processes: The larger an
organization, the more it tends having more levels of hierarchy. This
affects decision making – smaller organizations can react more quickly and
more efficiently to changing conditions.
* Feeling of ownership leads to higher job satisfaction: This one is
closely related to my last point: people in a smaller organization tend to
feel a much higher level of ownership over the outcomes of their work. As a
result, they’re more motivated (which, in most cases, will lead to better
* Stopping things that don’t work: Larger organizations tend to continue
projects although the outcomes of those projects are questionable. They can
simply afford it. Spinning off parts of the WMF would require the spin-offs
to justify their existence every single day. That seems to put a lot of
pressure on these spin-offs. However, the result would actually be good: if
an organization doesn’t continually deliver value to the ecosystem as a
whole, it will disappear. That’s better than a continued investment of
resources into projects that everybody knows don’t have the impact that
people expected them to have.
* More potential for innovation: Independent organizations have more
freedom to look at things from a fresh and different angle. While I was
still with WMF, everything needed to be done in MediaWiki. From the
perspective of the WMF that made sense. However, for what Wiki Ed needs to
accomplish, we believed that building our own software outside of MediaWiki
(but communicating with Wikipedia’s platform through OAuth) would be
better. Today, WMF is working on adapting Wiki Ed’s software to the needs
of a global community (see Edward’s email).
* Positive effects of competition: In a world, where organizations are more
independent, they won’t always choose a service provider within the
Wikimedia ecosystem. Instead they might decide to work with outside
contractors who are able to deliver better results in less time. As an
example, Wiki Ed worked with an outside contractor based in Seattle to
develop its dashboard. We could have partnered with WMF, but the project
might not have received the same level of attention. We’re happy with the
results, and now there’s one more company in the world that has a growing
understanding of Wikimedia’s requirements.
* Geographic diversity: if we decided to spin-off parts of WMF, we'd have
the opportunity to think about where these parts should be located. Given
the extremely high cost of living in the Bay Area and meager opportunities
for WMF to recruit people locally (for the many obvious reasons; e.g.
competing with Facebook and Google for talent is hard), some of the spun
off parts might be located in other regions of the world. This would make
the Wikimedia ecosystem more diverse than it is now and will most likely
== CON ==
* Coordination gets more difficult: With parts of the WMF getting spun off,
coordination between these parts will be more vital than ever. The risk is
that one organization doesn’t know what another organization is planning to
do / started executing / learned along the way. A solid process of
communicating things can solve this. However, this requires the willingness
of everybody to communicate early and often. And, in fairness, this is
already an existing problem in need of better solutions, given the large
number of chapters and affiliates WMF already has (and even within WMF
* Need for more EDs with unique profiles: With a greater number of
independent organizations in our ecosystem comes the need for more leaders
who are able to run a nonprofit in the Wikimedia world. We’ve already seen
that this isn’t easy to achieve. Nonprofits don’t pay as well as others,
EDs need to be comfortable with being yelled at on public lists, people
really need to understand our world. How would we be able to mitigate this?
(1) With more geographic diversity comes a larger pool of candidates
(actually, not everybody on this planet believes that it’s desirable to
live in San Francisco); (2) We should invest more in developing leadership
from within. WMF has a long track record of recruiting people “from
outside” for key positions. I believe we can generally do better in this
* More resources being spent on administrative stuff: If every new
organization creates its own finance / HR / communications department, more
resources in total will be spent on support work. That can either be
mitigated by creating specialized units which offer administrative/support
services to organizations within the ecosystem (at some point Garfield and
I discussed whether WMF should offer such services for Wiki Ed and/or for
affiliates) or by implementing these functions in a very lean way (that’s
actually the path Wiki Ed chose with e.g. having payroll done by an outside
contractor instead of doing it in-house).
* Smaller organizations might be weaker when it comes to partnerships: If
you’re the Wikimedia Foundation and you’d like to partner with, let’s say,
some Academy of Sciences, people at that institution will immediately
listen to you. Because you run one of the biggest websites on this planet.
If you’re a smaller organization, it’s more difficult to get someone else’s
attention. On the other hand, if we develop leadership from within, the
next generation of EDs may have existing connections to relevant partners
(as it was the case with Wiki Ed).
Now, do I believe splitting off parts of WMF and creating a “federation” of
organizations that deserves that name is better? Yes, I do. However, I
agree with Erik that this needs to be a “gradual process with lots of
opportunity to course-correct”. We have to let go of the idea that we’re
going to end up with an “ideal” organizational structure. We’ll have to
experiment and learn along the way. We might even have to reverse some of
our decisions in the future (which is ok). This all will take time, a good
amount of thinking, and everybody’s willingness to assume good faith. But
the recent developments at the Wikimedia Foundation have – at least in my
opinion – shown that thinking more thoroughly about the broader shapes of
our organizational structures is worth a shot.
I'm using this email address only for mailing lists. Personal emails to
this account will be lost.
On Thu, Mar 17, 2016 at 7:22 PM, Erik Moeller <eloquence(a)gmail.com> wrote:
Now that the dust has settled a bit, I would like to expand on an idea
that’s been touched on a few times (most recently, in an editorial by
William Beutler ): the notion that WMF might be a more effective
organization if it limited its own size in favor of focused spin-off
organizations and affiliates.
I was very much part of building the current WMF in terms of both size
and structure, but I also think recent events underscore the fragility
of the current model. WMF is still tiny compared with other tech
companies that operate popular websites, but it’s a vast organization
by Wikimedia movement standards. With nearly 300 staff  (beyond
even our ambitious 2015 strategic plan staffing numbers), it dwarfs
any other movement org.
I can see three potential benefits from a more federated model:
1) Resilience. If any one organization experiences a crisis, other
independent organizations suffer to a lesser degree than departments
within that organization.
2) Focus. Wikimedia’s mission is very broad, and an organization with
a clearly defined mandate is less likely to be pulled in many
different directions -- at every level.
3) Accountability. Within a less centralized federation, it is easier
to ensure that funding flows to those who do work the movement wants
them to do.
My experience is that growth tends to be self-reinforcing in budgetary
processes if there are now clear ceilings established. I think that’s
true in almost any organization. There’s always lots of work to do,
and new teams will discover new gaps and areas into which they would
like to expand. Hence, I would argue for the following:
a) To establish 150 as the provisional ceiling for Wikimedia movement
organizations. This is Dunbar’s number, and it has been used
(sometimes intentionally, sometimes organically) as a limiting number
for religious groups, military companies, corporate divisions, tax
offices, and other human endeavors.  This is very specifically
because it makes organizational units more manageable and
understandable for those who work there.
b) To slowly, gradually identify parts of the WMF which would benefit
from being spun off into independent organizations, and to launch such
spin-offs, narrowing WMF's focus in the process.
c) To aim to more clearly separate funding and evaluation
responsibilities from programmatic work within the movement -- whether
that work is keeping websites running, building software, or doing
Note that I'm not proposing a quick splintering, but rather a slow and
gradual process with lots of opportunity to course-correct.
More on these points below.
== Potential test case: MediaWiki Foundation ==
A "MediaWiki Foundation"  has been proposed a few times and I
suspect continues to have some currency within WMF. This org would not
be focused on all WMF-related development work, but specifically on
MediaWiki as software that has value to third parties. Its mission
could include hosting services as earned income (and potentially as an
extension of the Wikimedia movement’s mission).
MediaWiki is used today by numerous nonprofit and educational projects
that are aligned even with a narrow view on Wikimedia’s mission.
Examples include Appropedia, OpenWetWare, WikiEducator, W3C’s
WebPlatform, Hesperian Health Guides, and too many notable open source
projects to list.
Among commercial users, it has lost much ground to other software like
Confluence, but it remains, in my view, the most viable platform for
large, open, collaborative communities. Yet it’s a poorly supported
option: many of the above wikis are outdated, and maintaining a
MediaWiki install is generally more work than it needs to be.
Building a healthy third party ecosystem will have obvious benefits
for the world, and for existing Wikimedia work as well. It may also
create a proving ground for experimental technology.
Which work that WMF is currently doing would be part of an MWF’s
mandate? I don’t know; I could imagine that it could include aspects
like Vagrant, or even shared responsibility for MediaWiki core and
== The Wiki Education Foundation precedent ==
It’s worth noting that this spin-off model has been tried once before.
The Wiki Education Foundation is an example of an organization that
was created by volunteers doing work in this programmatic space in
partnership with staff of the Education Program at WMF, who left to
join the new org. It is now financially independent, building its own
relationships with funders that WMF has never worked with, and
achieving impact at unprecedented scale.
LiAnna Davis, who is today the Director of Program Support at Wiki Ed,
wrote a detailed response to William’s blog post, which I think is
worth quoting in full :
I worked for the WMF for nearly four years and have worked for the
spun-off Wiki Education Foundation for the last two, and I strongly
support the idea of spinning off more parts of WMF into independent
nonprofits like ours.
As you noted, Wiki Ed is a test case for your proposal, so for readers
who don’t know our history: We started in 2010 as a pilot program
(called the Public Policy Initiative) within WMF, funded by a
restricted grant, to support university professors in the U.S. who
wanted to assign their students to edit Wikipedia as a class
assignment. The pilot showed the idea was successful, and so we
started piloting it in countries as part of the Catalyst project (Arab
World, Brazil, and India).
The U.S. program had lingered at WMF without any real organizational
support because the U.S. wasn’t a target region. WMF leadership saw
its potential, however, and formed a volunteer Working Group of
Wikipedians and academics who created the structure of the
organization that became the Wiki Education Foundation in 2013. WMF
gave us a small start-up grant to get us going, and provided fiscal
sponsorship for us until our 501(c)3 status came through (and we could
fundraise on our own).
Today, we’re an independent organization, not funded by WMF, and we’ve
scaled the impact of our programs incredibly. We’re supporting three
times as many students, we’ve developed our own technology to support
our programmatic work, and our students are busy addressing content
gaps in academic areas on Wikipedia.
So why are we so successful? There are a lot of factors, but there’s
one I want to highlight here, because I think it’s a clear difference
between when we were at WMF and our current work at Wiki Ed. We have
one, very clear mission: We create mutually beneficial ties between
Wikipedia and academia in the U.S. and Canada.
The WMF mission is inspiring — but it’s really broad, just like our
movement is. When we were doing this same project at WMF, I’d struggle
to just focus on the Education Program and ignore the rest of the
mission. Whenever I interacted with people outside the foundation (and
I did so a lot), people would come to me with ideas to further WMF’s
mission that weren’t in my program’s boundaries. I’d spend time trying
to help, because I believed in the mission and wanted to help it
along. I’m not the only one: I would see this idealism and commitment
to the mission repeatedly among my colleagues at WMF. I still see it
from the current WMF staff. They’re all there because they believe in
the mission. They want to help, and it’s really hard to not try to
help with everything, because you can see so many different facets of
helping that mission.
Essentially, with a mission as broad as WMF’s, it’s hard for staff to
keep a narrow focus. *Everything* can seem mission-related. When your
mission is as narrow as Wiki Ed’s, it’s easier to find your focus and
keep your attention on developing one area well. This is a key
strength of independent organizations — independent, narrower missions
keep staff focused and more productive on achieving their small part
of the overall Wikimedia mission.
I strongly support more discussion about spinning off other parts of
WMF into independent organizations.
== A "Movement Association"? ==
A more radical suggestion would be to spin off work on grantmaking and
evaluation. This isn’t trivial -- there are legitimate arguments to
keep this work close to other community-facing work WMF is doing. But
there are undeniable benefits in greater separation.
When it comes to large annual plan grants, much has been done to
ensure that the FDC can operate as an independent body and evaluate
each plan on its merits. Ultimately, however, the decision rests with
the WMF, which has a much better understanding of its own programs
(through the direct relationship with its ED) than of those of
Similarly, while WMF has done a fair bit to provide self-service
evaluation tools to the movement at large, it’s not clear that its
work is always held to the same standard as everyone else’s. A WMF
grantee must very publicly report results and success metrics; WMF
attempts to do so as a matter of course, but it is not accountable to
another organization for failing to do so.
Finally, as was discussed here a lot in recent weeks, WMF itself has
no clear accountability to the movement. The Board elections are
advisory in nature. There is no membership. Non-elected seats are
filled by the Board with little visibility. There is a semi-permanent
If grantmaking and evaluation responsibilities were increasingly
shifted to a "Wikimedia Movement Association", this could gradually
allow for true accountability to the movement in the form of
membership and democratic, movement-wide decisions to make funding
allocations on the basis of evaluation reports (through committees or
This may also make the endowment a more compelling proposition than it
is today. Yes, keeping Wikimedia’s sites operational indefinitely is a
very worthwhile goal. But what if the endowment ultimately also helped
to support global, federated work towards Wikimedia’s vision? What if
all affiliates -- indeed the whole movement -- were excited and
motivated to help grow it?
== Where to go from here? ==
There are lots of open questions in all of this. Should all site-wide
fundraising remain inside WMF, for example, with funds being
transferred to a movement entity? What’s the dividing line between
"development for third parties" (MWF) and "development for
(WMF)? How would staff transition to new organizations? Where should
those organizations be based? Should they be distributed, have
An important thing to remember here (a lesson I’ve had to learn
painfully) is that big changes are best made in small steps, with room
for trial and error.
Implementing this strategy is, I think, a matter of first committing
to it as an idea, and then creating coherent proposals for each step,
publicly with broad input. First, if there is support for the general
idea, I would recommend kicking it around: Are these the right kinds
of spin-offs? What are the risks and how should existing affiliates be
involved in the process? And so on.
The fact that WMF has just experienced a major organizational crisis
should not itself fill us with pessimism and despair. But we also
shouldn’t ignore it. We must learn from it and do what reason tells us
-- and in my view that is to build a more resilient _federation_ of
organizations than what we have today.
== Notes ==
 Our branding is confusing beyond repair. I don't think there's an
easy fix here, and we should just embrace our nutty nomenclature
(Wikimedia/MediaWiki/Wikipedia) at this point.
Wikimedia-l mailing list, guidelines at:
New messages to: Wikimedia-l(a)lists.wikimedia.org