Support the People of Wikipedia

During my tenure as Vice President of Engineering at Wikimedia Foundation, before anyone met with the Knight Foundation, I participated in executive off-sites with all execs of WMF.  This included bringing in professional coaches to coordinate the sessions as well as individual strategy coaching efforts provided by the Executive Director, and her personal coach, with whom I met several times. By my observations, the CEO from The Leadership Circle was master of ceremonies at most of these sessions based on our Executive Director’s guidance of topics.

I believed, at the time, that these sessions were intended to create a strategy to help Wikipedia by making sure all WMF groups are aligned and executing towards a single goal.  Instead, our ED coordinated these sessions with the general topic: What silver bullet-type software feature can we create to save Wikipedia?

These are two very different things:

The first one (creating a strategy to help Wikipedia my making sure all WMF groups are aligned and executing towards a goal) is a collaborative effort that, as a consequence of what is actually causing any decline of Wikipedia may not actually be software or feature related, requires a direct balancing of resources.  Specifically:  If WMF were to behave as a software engineering company because it had a very specific software goal that might help Wikipedia, it will invest in engineering.

However, if the goal were to organize a movement, WMF might organize itself as a media superpower and begin investing in communication efforts and media specialists to leverage open source communication channels and community collaboration in the most transparent way possible. In these sessions, I expressed that we must choose the path of a media superpower, because trust is the superpower of Wikipedia. We all say we can’t trust it, but the truth is we all go to it because we do trust it enough to love it and challenge it to produce better answers constantly, by criticizing it and contributing to it, often several times a day.

I believe it is almost impossible to generate an idea for a software solution to save Wikipedia in a closed room in discussion with only a handful of people.   Also it restricts any strategy to one that involves running an engineering team de facto.  It assumes that whatever we are going to do, it’s going to be in software.  I’ve seen this before at Mozilla where a small set of non-technical contributors attempt to ideate a software feature that would save Firefox from Chrome. I call this behavior:

Homeopathic Software Engineering:  Attempting to save a massive complex software product in the marketplace from a competing equivalent product by adding an insubstantial feature; making an insignificant change touting dramatic improvements based on little or no evidence any such changes will help; often attempted out of fear or in light of another, usually better, option that Executive Leadership will avoid with extreme prejudice because they might be wrong.

The ED demanded a Silver Bullet software solution. But this was bigger than any software engineering effort.   This is a global-community-coordination effort to stop the decline of Wikipedia.  It’s very different.

There were many of these strategy sessions.  The team became frustrated after our ED rejected acting on every effort proposed.  The ED requested constant modifications to any proposed plan that stripped it of any potential for actually affecting any change.  The teams struggled to come up with something that the ED wanted.  It never materialized because it’s a fruitless effort.  It frustrated the executives as there was no end in sight in determining any strategy whatsoever.

The Executive Director, who was my boss, began putting tons of pressure on me to come up with something she could use as a strategy. The organization was clearly growing confused and frustrated by the lack of clarity.

I received a request from the ED to meet with her personal executive coach and mentor assigned with the goal to come up with a silver bullet plan for her. I met with her mentor for at least two sessions over a couple of months, and he asked me to answer one question:  What would have to happen for Wikipedia to win?  He asked me to write it down and we’d review it together with the ED.

I really liked his question, and I believed it an opportunity for me to explore and write down my ideas in a very concrete way.  Of course, I argued with him over what the word ‘win’ means. I simply took it as a challenge to write down, explicitly, the words describing what I think must happen for Wikipedia to be saved—not ‘win’.

I came back two weeks later with a proposed plan.  I called it The Stand.   It consisted of these things in order:

  1. Demonstrate WMF is competent at what we do to support Wikipedia.  i.e., make sure we can do what we say we can do, and that we can determine conclusively if Wikipedia is dying or not, because we must be sure about this one fact.  It’s fundamental to the survival of Wikipedia.
  2. Publicly become a loudspeaker for what is happening to Wikipedia:  What must change for it to continue to survive?;  Give hard data and evidence showing how commercial consumers of Wikipedia are both improving and hurting Wikipedia by way of API usage, Search results integration, Impact of Knowledge Vault on Wikipedia, and intermediation of editors by way of search results which keeps people from editing Wikipedia over time, etc.
  3. Since intermediation by search engines are the most obvious threat to Wikipedia, it’s the one threat the entire organization believes is real, immediately begin improving our Search section of Wikipedia to be just good enough to provide a transparent search mechanism for Wikipedia–in partnership with other open search projects like Open Street Maps, etc.–in preparation for the potential event that Wikipedia should need become the default Search provider for Wikipedians (and the World) .  If we do not achieve Wikipedia growth by influencing corporations, governments, and any other organization by showing the world how they are affecting Wikipedia, using public data in a way that the public would understand, that intermediates Wikipedia from its users, then we must be able to provide a transparent search feature that can be used as default by those who want transparently curated search results from Wikipedia.

IMPORTANT:  The point of becoming a loudspeaker–a signal to the world that cannot be ignored–is so that WMF can rally with other open source movements to affect the change needed in order to save Wikipedia.  WMF can’t do it alone, and it must do it publicly from the beginning in order to gain the support and momentum for any significant technology/software/open source engineering challenge we might discover and face.  It’s important to know that everything was planned to be done in public from the very beginning, including discussing the plan, and what the data reveals, before making any strategic decisions.

The Executive Director expressed fear of the plan because she thought Google would squash Wikipedia if we talked about how they, and every other corporation feeding off of Wikipedia, were not only consuming, but influencing traffic to/from Wikipedia and potentially systematically killing it.

The first task she gave me and the executive team was:  Create a business plan that shows that The Stand is a viable effort or else we can’t move forward.  But, there’s a catch, because I can’t create a business plan that says:  If Google, Microsoft, and Apple are killing Wikipedia by intermediating our users in search results, go build a loudspeaker and prepare to replace search.  If Google’s killing Wikipedia, that’s an obvious financial reason to charge major corporations for access, but she already said we’re not allowed to talk about that, and she shut us down in all strategy conversations that went there.  The only thing we were allowed to do is make a compelling argument for a search project to be funded somehow. If I could create a business plan that shows that it’s a good idea, that it will make money, or is financially reasonable for WMF somehow, then she would agree to move forward.

I insisted that I believed it was a bad idea to create a business plan because we’re not about making money, and I really don’t think that Search is going to be the thing that saves Wikipedia.  I believe what will save Wikipedia is the Loudspeaker as its purpose is to show what, exactly, is happening to Wikipedia, and who has the influence, power, and wherewithal to save it.  It isn’t about creating a killer feature.  Search is the just tool that we may have to use in order to gain the leverage we need to guarantee Wikipedia growth, but we’ll only be able to do so in the presence of a vast open public community of projects willing to create and support such a massive project as search in Wikipedia.

I created a business plan with the team.   I was frustrated but determined to overcome this fake hurdle.  I worked with all of the other execs to create the proposal based on my ideas.   It was the Head of Fundraising who came up with the idea of using her team to write up a restricted grant proposal as a form of a business plan that would make sense to Wikimedia.  After all, we don’t really have business plans, we have proposals that take into consideration budget, etc.  And, we don’t actually have to propose anything to any potential backer because it’s just an exercise.  Her team took my plan, interviewed me exhaustively, and worked up a really great argument for why we should improve Wikipedia search in the form of a grant proposal.   I suspect that if this exercise of creating a grant proposal to clear a hurdle for our Executive Director had never happened, we’d never have seen the Knowledge Engine Knight deal at all.  Later, I suspect our Executive Director saw this as an opportunity to cherry pick an almost-viable tech idea she thought might be successful just long enough to maintain a positive-looking tenure at WMF and also give her time to stock the board with members who will guarantee her own success when she decides to move on in the corporate world.

The Executive Director supported the grant proposal idea, and the next thing I know, I’m being asked to present the proposal to the Knight Foundation.  The reason, I was told, was to test the waters for the bigger plan of Loudspeaker and building a search coalition with other projects.  They wanted me to present my plan to the head of the Knight Foundation to see how they would react.  I expressed that any small effort to fund something as massive as search would be pointless, but the ED and Head of Fundraising insisted that we could use this opportunity to reach out to other potential donors who might have an interest in helping out, should the Knight Foundation be interested, as they could connect us directly to those who would want to see transparent search implemented on Wikipedia.

They convinced me.  I didn’t see any reason to stop it from happening as I felt any conversation about what will save Wikipedia should be allowed to continue.  It didn’t seem like a threat at the time.  Also, it seemed improbable that we’d pursue a restricted grant to fund search because the amount of money it would take is far beyond what they would be willing to donate (tens of millions of dollars is what I recommended on the proposal as cost for search improvements—just to get us up to “good enough” was well over six million dollars by back-of-the-hand calculations of resource costs).

I presented the limited grant proposal to the Knight Foundation and I also emphasized that we’d need to build an open coalition to do anything meaningful with search on Wikipedia. The Knight Foundation was interested improving independent journalism in the United States.  I stated that everything would be done in the public from the very beginning.

The Knight Foundation was very supportive and asked a lot of great questions. I did not know that I would not be around to be a part of future meetings with the Knight Foundation.

I asked my boss, the Executive Director, to give me the go ahead to start making changes in my department that would restructure us into an organization that would enable us to do The Stand.  Again, I expressed that it would take more than an engineering effort to do something like I’m suggesting, and that it would take a fundamental restructuring of the entire organization to focus on the single effort of guaranteeing the survival of Wikipedia by way of leveraging community and other open source projects.

Before the Knight Foundation meeting, and outside of my purview, the Executive Director hired a new executive to lead the search effort to report to me and she gave me instructions to find a place for him in the organization. This was an impossible situation.

This act of cronyism had a profound negative impact on my team.  It’s a unilateral act which displaced a key member of my staff, who at the time was helping me guarantee WMF Engineering’s overall success, out of the running for a planned leadership promotion.  It killed the morale of the team dead.  They lost faith in my ability to lead.  I could no longer keep members of my staff on a track to success in their eyes.  The Executive Director demonstrated that she would step on me if she wanted to and that I didn’t have the power to stop it.

Before presenting to the Knight Foundation, I began asking the Executive Director to allow me to present my ideas, The Stand, to the board of trustees.  I asked repeatedly for weeks:  Please let me present this to the board as they must know what is going on.  She instructed me to not talk to anyone on the board about my concerns or ideas.

Next she insisted that I continue to redesign the org around goals that did not include any effort other than Search with the new executive she hired in charge, and that I should do a reorg with the expectation that I would become both the head of product and still lead engineering as she wanted to move forward with our plans.   She indicated that I should plan on approximately 35% growth in my teams and move forward with a reorg based on a basic implementation of my plan.  She also indicated that I should not discuss The Stand, again which is:

  1.  Loudspeaker
  2.  Show The World Who’s Killing Wikipedia, and
  3.  Start Improving Search Only Because We Must

..with anyone any more.  Specifically, she said that I’m not to mention the reasons behind search (which had been stripped from being a world-wide open source collaborative effort to one that is powered by a small search team led by this new executive), which was to be as part of the plan to actually save Wikipedia overall.  I couldn’t believe it.   It was clear that she was willing to bring in people without asking and that means she’d likely get rid of me just as fast.  So, I knew I was walking on thin ice at that point.  Especially since I’m no longer allowed to discuss my ideas with anyone.  It was a horrible place for me to be as a leader!

But, she insisted I move forward.

So, working with my teams I redesigned the organization around Read, Edit, and Search.  I communicated to the rest of the executives that I believed we must re-balance our budget and resource allocations for the entire organization in order to build the loudspeaker first, and that would require potentially reducing my budget in order to grow the Communications group dramatically, as well as the community groups at WMF, because that is exactly the place we need to invest first.  In parallel with investing in specific, targeted engineering efforts in data science in order to determine, concretely, if Wikipedia is growing or not, we must also begin coalition building and the largest awareness campaign Wikipedia has ever seen.

And then came a huge bombshell:  The Executive Director and a new outside contractor acting as COO informed us that we don’t have the budget to do what we wanted to do, and now that we’ve reorganized, we must tell the teams that they can’t hire those people that they needed in order to move forward with their proposed goals.  The Executive Director demanded that we stop everything and do a budget based on the current goals of Reading, Editing, and Search.  I insisted that this would leave us in a horrible position of having to explain that we cannot do the things we said we were going to do, most important: Loudspeaker.  But management was crushed.  We were totally blindsided by the budget being cut and then we’re being forced to create a budget for something we didn’t believe in.  This sent the executives into a state of shock and confusion.

We completed the reorg because we had to.  The teams had been in disarray for too long.  Since I couldn’t discuss the overall strategy in any meaningful way due to the restrictions placed on me by the Executive Director, the only option left for me was to immediately begin publicly orchestrating our open source movements and creating goals in public that the Executive Director couldn’t force me to change because they were collaboratively created in public, and I began first in IRC and was going to move to the mailing lists immediately as soon as I bolstered the right leaders directly via IMs.

The Executive Director immediately recognized that I was moving forward publicly.  She was in IRC and even replied, ableit she’d never actually follow through as I identified later, with support when I stated in public that we’d be creating our goals in public, and it’s up to us as community leaders to make our goals collaboratively.   Right after she read this on IRC, she called me into her office and told me that I had two weeks to come up with another title for myself, that was no longer VP of Engineering or Head of Product, and that if she liked what I produced, she would let me keep working at WMF.

Of course, I was in complete shock at such a demand.

She explained how she felt it would look bad for her if she fired me right away, so she wanted to give me a chance to come up with something to help her save my job.  She also instructed me that I wasn’t to discuss this with anyone.  I immediately refused to comply with her command because I do not believe it honest to create a new job for myself in a vacuum from my team, and I see no reason why such an effort would be successful.  I only see my team being confused and diminished and it in no way helps the effort.

The reason she gave me for removing me was that “my team no longer trusted me.”  I demanded specifics.  What, exactly, did they not trust me about?   She wouldn’t say.  I asked her who I could talk to about them not trusting me and why, and she indicated that I’m not allowed to discuss it with anyone even the executives who are my peers.  I told her I would not comply. She again parroted that I should come up with another job description and title for myself, and she would review it when she returned.  And that is how we ended the conversation.  She went on her business trip for two weeks.

I kept my word and did the only thing I thought an honest person would do:  I immediately began discussing my departure with my direct leadership staff.  I told them that I was going to be removed in two weeks.  I did not have a choice.  And I told them that I thought I would likely be removed before those two weeks were up as this, for our Executive Director, is really an untenable and explosive situation, and frankly she’s demonstrating remarkably poor judgement in putting me in such a position.  I told them them I’ve been asked to not discuss this with them and that I could not comply with that command as I believed it unethical.

The next day, with security and outside legal counsel observing, I was walked out of the building because, in the Executive Director’s words:  “Damon, you are doing damage to the Wikimedia Foundation.  You are hereby put on two weeks paid leave.  Effectively immediately.”  I demanded to know what damage I’d caused to WMF, and she said that by talking to my team and indicating my departure, I did irreparable damage to the Wikimedia Foundation.

When I reminded her of her statement that I had two weeks remaining in my role and that I refused to not discuss that fact with my team, to her directly in our conversation, she did something unbelievable.

She denied ever having the conversation with me.

Right there in front of everyone.  In front of lawyers.  And there was nothing I could do to prove otherwise as it was in a one-on-one that she made these demands of me. We were alone.   I immediately claimed that it was a lie on the spot.  I got up out of my chair and demanded she tell the truth.  She refused.  She again said she never had that conversation with me.  And this incident of me making such a claim just proves that I can’t be trusted.

I said, or rather, I yelled, because I was angry at such treachery, “That’s a damn lie!”  And I walked out without signing anything.

Five or so days later, I received a certified letter in the mail saying I was fired and offered a severance/silence agreement that said I was not allowed to criticize WMF in exchange for 60 days of pay.  Knowing that it would take me longer than two months to recover, I asked for more, and I signed because, I didn’t take this job to make money.  I took the job because I wanted to help.

Since I’m not rich, and I’ve never achieved any significant financial success , I’m living hand to mouth with my family of six.  Even then, I knew I couldn’t afford a six month job loss.  Also, I didn’t want to hurt WMF.  I didn’t want a scandal.  So, I just signed.  I decided that we’d make the best of things and I’d just get another job as quickly as possible and try to put this behind us.  I asked for four months and they agreed immediately.  It took me five months to find another job when I lost it before, so I thought I could pull it off. I was wrong.

It’s been almost three years now.  By working at WMF, it feels as if I destroyed my career as a VP of Engineering anywhere.  From this, I’m attempting to transform myself into something better and more effective.  I’ve become a better person in the process, I feel.

After I left, they must have continued to pursue the idea with Knight.   The deal was done after I left, obviously.  But these ideas needed to surface and be discussed publicly, and James Heilman changed things for the better.  He sniffed it out right away as soon as he arrived:   What’s this effort around Search?  Shouldn’t we talk about this publicly?  Is this the best way to do this?

He was right in doing so.  Millions are potentially wasted on no real effort at all.  Suppressing actual ideas that could save Wikipedia and obfuscation of WMF’s efforts has caused a loss of time, money, and most important: Amazing people. That is tragic, but we shouldn’t let this deter us from the real goal: Supporting people’s freedom of speech, freedom from violence, and a right to thrive. If we can do that, I can live with Wikipedia and its success.

The real story here is that corporations, even not-for-profit efforts like Wikimedia Foundation, are autocratic and fascist entities: One person is a dictator who can be completely corrupt without consequences. Regardless of designation, if money is involved and there’s a single person or group with all of the authority, profits will prevail over people.

It’s time to do away with corporate power hierarchies and replace them with democratic structures where everyone in the company has a vote and nobody can be fired or hired without a public discussion. I no longer participate in private corporate efforts. I find them dishonest and unable to demonstrate accountability.  Most are downright abusive. I hope I can demonstrate to others to do the same.

One thought on “Support the People of Wikipedia

  1. Pingback: Wikipedia Search | Ministry for the Poor and Vulnerable

Comments are closed.