By Andreas Kolbe
The other day, I was fortunate enough to be treated to a fundraising banner on Wikipedia:
DEAR WIKIPEDIA READERS: To protect our independence, we’ll never run ads. We survive on donations averaging about £10. Now is the time we ask. If everyone reading this right now gave the price of buying a programmer a coffee, our fundraiser would be over within an hour. We’re a small non-profit with costs of a top 5 website: servers, staff and programs. Wikipedia is something special. It is like a library or a public park. A temple for the mind where we can all go to think and learn. If Wikipedia is useful to you, take one minute to keep it online and ad-free another year. Thank you.
I have just donated to #Wikipedia. Help keep it free! #keepitfree
Donations link added, of course.
There’s just one problem here: the Wikimedia Foundation (WMF) – the non-profit that runs Wikipedia and other crowdsourced projects such as Wikimedia Commons and Wiktionary – is wealthy enough today to keep Wikipedia “online and ad-free” for about a decade.
So why does it want even more of your money?
1,000% growth in revenue
Under Sue Gardner’s leadership (2007–2014), the Wikimedia Foundation’s revenue skyrocketed, based on fundraising banners designed to solicit small donations from a very large number of Wikipedia readers.
The Wikimedia Foundation does extensive testing of its fundraising banners each year to establish which banner wordings, which text highlights etc. are most effective at converting page views into dollars. The banner I saw was, I presume, one of those tests. “The price of buying a programmer a coffee” makes a good soundbite. I shouldn’t be surprised if we’re all invited to buy one of their programmers a coffee once the annual fundraiser kicks off in earnest later this year.
People tend to respond with a slight tinge of panic to the prospect that Wikipedia might go offline. So it’s virtually certain that millions will once more reach for their credit cards and pay up.
A look at the Wikimedia Foundation’s financial records however quickly shows that its fundraising drives have little to do with keeping Wikipedia online. Keeping Wikipedia “online and ad-free” is simply an effective fundraising slogan that’s survived WMF fundraisers’ Darwinian selection process.
$45 million in net assets
At the close of its 2012/2013 financial year (ending June 30), the WMF reported $45 million in net assets. (This included $22.2 million in cash and cash equivalents, plus $17.6 million in investments. That year alone, revenue exceeded expenditure by $13 million.) I would estimate that a positive year-end balance of $45 million to $50 million is enough to keep Wikipedia “online and ad-free” for about ten years, based on the simple fact that in 2007/2008, when Wikipedia was already a top-10 website – with a rather larger editing community than today – the WMF got by on annual donations totalling $5 million, a tenth of what it took this year, and an annual expenditure of $3.5 million.
The table above reflects the most recent financial statements available, but in a recent video, a WMF staff member announced that the Foundation had taken another $50.5 million in the period from July 2013 to June 2014, exceeding its projected spending of $50 million for 2013/2014. In fact the WMF has for years now had a steady pattern of spending below plan: in 2012/2013 it underspent by $3.6 million (9%). The surplus money was added to its reserve.
As the chart shows, Wikimedia Foundation revenue, expenses and assets rose by about 1,000% over a period of five years.
Where is all this money going, if it isn’t needed to keep Wikipedia online and ad-free?
Improving the reliability of Wikipedia costs money, doesn’t it? Well, no, actually.
Many donors may be under the mistaken impression that the Wikimedia Foundation must be using its money to improve the reliability of Wikipedia content, say by paying for academic expert review, or employing staff to root out vandalism.
This is not so. The WMF does not write any Wikipedia content, nor does it employ experts or other staff to review it. In fact, the WMF does not even measure content quality – their staff admit freely that after 13 years in the education business, they still have no idea how to do it.
The WMF is quite happy to leave content generation and quality control in the hands of its unpaid and largely anonymous volunteers, magnanimously described as “editors”. (In Wikipedia parlance, anyone who clicks “Edit” on a Wikipedia page and saves a change thereby becomes a Wikipedia “editor”. See?) The results are a bit patchy, but who cares if it’s good enough for Wikipedia to be a top-10 website?
What the Foundation does measure and report is the number of volunteer editors (they’re worried that numbers have been in decline since 2007), the number of articles (steadily increasing), the number of article edits (the more, the better – even if performed by non-human “bots”), page views (dropping), the site’s Alexa rank (currently no. 6), and the amount of donations flowing into its coffers (steadily increasing).
“Log-rolling and self-dealing”
Instead of funding measures to improve content quality, income from donations has supported a ballooning bureaucracy made up to a significant extent of former volunteer contributors, both at the Wikimedia Foundation itself and in its affiliated national and regional chapters, whose levels of paid staff have risen along with those of the WMF itself.
I believe that currently, too large a proportion of the movement’s money is being spent by the chapters. The value in the Wikimedia projects is primarily created by individual editors: individuals create the value for readers, which results in those readers donating money to the movement. We have over 40 Wikimedia organizations today, 12 of whom received funding allocations through the FDC last year. Of the US$5.65 million WMF gave out in grants last year, 89% or US$5.04 million were to affiliate entities, with US$4.71 million (83% of the total grants) to these 12 entities for their annual plans. I am not sure that the additional value created by movement entities such as chapters justifies the financial cost.
I believe the FDC [Funds Dissemination Committee] process, dominated by fund-seekers, does not as currently constructed offer sufficient protection against log-rolling, self-dealing, and other corrupt practices.
Does this sound like an organisation struggling to keep Wikipedia online and ad-free? No. It sounds like an organisation struggling to keep a handle on where all the money is going.
Compared to the early days of Wikipedia, the influx of money has created a two-tier society: those who were lucky enough to get a paid job with the Foundation or one of its chapters, and those who keep toiling away unpaid in Wikipedia – reverting vandalism, eliminating puff-pieces and hatchet jobs, checking for copyright violations, writing and updating content. Behind the scenes of Wikiland, there is recurrent grumbling from unpaid volunteers about misspent funds and about WMF and chapter staff living the jet-setting life, with travel budgets funded from donations.
A miserable cost-benefit ratio
In 2005, the Wikimedia Foundation had 1 employee, and the site’s software was written and maintained by volunteers. In fact, in 2005 Jimmy Wales proudly told a TED audience how little it cost to run Wikipedia:
“So, we’re doing around 1.4 billion page views monthly. So, it’s really gotten to be a huge thing. And everything is managed by the volunteers and the total monthly cost for our bandwidth is about 5,000 dollars, and that’s essentially our main cost. We could actually do without the employee … We actually hired Brion [Vibber] because he was working part-time for two years and full-time at Wikipedia so we actually hired him so he could get a life and go to the movies sometimes.”
Today, the Wikimedia Foundation attracts 21 billion page views a month – i.e. 15 times as much – but even 15 times the $5,000 a month Wales mentioned would only be $75,000 a month, or $900,000 a year; and that is without allowing for economies of scale, and the fact that bandwidth has become cheaper since 2005.
At the time of writing, the Foundation’s paid staff and contractors page lists 215 people (not counting various mascots). They comprise 2 people in the Office of the Executive Director Lila Tretikov, 18 in Grantmaking, 17 in Fundraising, 14 in Legal and Community Advocacy, 4 in Communications, 17 in Finance and Administration, and 11 in Human Resources. The remaining 132 – almost two-thirds of the total – all work in Deputy Director Erik Möller’s Engineering and Product Development department.
WMF Deputy Director and Head of Engineering and Product Development, Erik Möller.
That department, whose direct and indirect costs account for about one-half of WMF expenses, is currently not popular with the unpaid volunteer community, given a series of high-profile failures to deliver functional software.
The VisualEditor (VE), announced with great fanfare in the press last year and described as “epically important” by Jimmy Wales, was a WYSIWYG text editor that was supposed to revolutionise Wikipedia, making its articles as easy to edit as a word processor document and bypassing the site’s complicated and idiosyncratic mark-up language. When it was finally launched, years behind schedule, the volunteer community hated it, calling the new feature “buggy” and complaining that it broke articles (which it did, in many unexpected ways, e.g. by inserting chess pawn ♙ characters into Wikipedia articles) – errors that unpaid volunteers then had to fix manually.
In addition, the VisualEditor was only able to handle tasks that were relatively easy already, meaning that users still had to learn Wikipedia’s mark-up language for advanced functions like tables and referencing. Eventually, an administrator went over the Foundation’s heads and disabled the code that made the new feature the default editor. The Foundation backed down.
Wikipedia’s annus horribilis
This year, things have only gotten worse. In August, the WMF faced an outright revolt from the English and German Wikipedias’ volunteer communities over another unloved new feature, the Media Viewer. A volunteer administrator on the English Wikipedia tried to disable the feature, much as the VisualEditor had been disabled the year before. In response, WMF Deputy Director and Head of Engineering and Product Development Erik Möller threatened to remove the volunteer’s administrator privileges – and was hauled in front of Wikipedia’s Arbitration Committee for throwing his weight around. (The case is currently suspended.)
Faced with a similar admin revolt in the German Wikipedia, the WMF created a new access right, Superprotect, to put the relevant configuration file out of volunteer administrators’ reach. The result was that Möller’s own user account on the German Wikipedia was blocked. This was followed by an overwhelming vote of no confidence from the German volunteer community, with the highest voter participation in the project’s history, and an open letter of protest to the Wikimedia Foundation from the international volunteer community that attracted over 900 signatures – again an unprecedented number. Not even the celebrated January 2012 Wikipedia black-out that stopped proposed US anti-piracy legislation in its tracks attracted as many support votes from Wikipedians.
A cartoon drawn by a Wikipedian, satirising the “Superprotect” controversy.
Jimmy Wales, beset by angry volunteers on his talk page, acknowledged that some of the criticism was justified. Confronted on his talk page with a quote from a tech writer who’d opined that
“The Foundation has a miserable cost/benefit ratio and for years now has spent millions on software development without producing anything that actually works; the feeling is that the whole operation is held together with the goodwill of its volunteers and the more stupid Foundation managers are seriously hacking them off”,
“Other than the extreme nature of the comment (‘without producing ANYTHING’ is too strong) why do you think I would disagree with that? This is precisely the point of the new CEO and new direction – to radically improve the software development process. That statement, while too strong, is indeed an accurate depiction of what has gone wrong. I’ve been frustrated as well about the endless controversies about the rollout of inadequate software not developed with sufficient community consultation and without proper incremental rollout to catch showstopping bugs.”
But the next WMF software project, Flow, a planned Facebook-style revamping of Wikipedia’s discussion pages that has been in development for some time, is already mired in controversy, with some volunteer editors threatening to strike if it ever sees the light of day. Many of them feel that a Facebook discussion format aimed primarily at mobile users will be too inflexible to serve their needs, and that the Foundation is turning a deaf ear to their concerns.
It ain’t what it used to be
Now remember that Wikipedia worked perfectly fine when there weren’t any paid staff programmers at all. The original version of the MediaWiki software that Wikipedia runs on was written by Magnus Manske, a German-born volunteer. He did it while on holiday:
Why did he do it? He says he saw the need, thought he could help, and wanted to learn a new programming language called PHP that he hadn’t used before.
Remarkably, it took him just two weeks. He submitted his program to Jimmy Wales and the rest of the Wikipedia team and they spent another six months testing and refining his software. Almost a year after Wikipedia first launched […] the site migrated across to Magnus’ free, open source software – now known as Mediawiki.
That was the original spirit of Wikipedia. People chipped in with volunteer labour whenever something needed to be done. But when the fundraising drives conducted under Sue Gardner’s leadership proved a roaring success, there was a lot of money accumulating suddenly. And wherever a lot of money accumulates, pockets eager to be filled magically appear.
From the very beginning, Wikipedia’s volunteer community was a tech-centric lot, composed mainly of young programming geeks rather than educators. About half of all Wikipedians are school or university students; less than a third are in full-time employment. It was quite natural for many of them to jump at an opportunity to turn their unpaid Wikipedia hobby into a paying job.
When all you have is a hammer, every problem looks like a nail. And when you’re an amateur programmer, every problem – even in an educational enterprise – looks like a software problem. So it shouldn’t come as a surprise that as the Wikimedia Foundation’s staff grew, this growth was focused above all on acquiring more tech staff, and that many of the new management, software engineering and support positions created in this process were filled with Wikipedia insiders who displayed fervent faith in the Wikipedia idea, combined with an often total lack of relevant industry experience.
Lila Tretikov, the Wikimedia Foundation’s new Executive Director.
Lila Tretikov, who took over from Sue Gardner as the new WMF Executive Director this summer, has her work cut out to make her Engineering and Product Development staff deliver work that will stand up to muster.
There is no doubt that the Wikipedia user interface could do with updating. Women in particular, who are as underrepresented in Wikipedia as ever (estimates range from 8.5% to 16%), might find a more aesthetic and contemporary user experience more appealing. Wikipedia runs the risk that readers will vote with their feet, flocking to third-party sites like WikiWand to read Wikipedia content, where they will no longer see the WMF’s fundraising banners.
Despite employing now over 130 people and spending millions of dollars per year in product development under Möller, Wikipedia has become a creaky, aging, and visually backward web platform, lagging far behind other online products with equal or fewer development resources. And indeed, they seem with every new release to be slipping further behind, as well as creating controversy for their users. Tretikov is well aware of these challenges, and unlike a good number of her staff, she is not a Wikipedian but someone with a proven track record in the software industry.
What’s it all worth?
As one of the world’s top-10 websites, Wikipedia’s potential market worth is in the order of billions. But it also has a Creative Commons licence that allows unlimited commercial re-use, which means that the value generated by its army of unpaid volunteers benefits a multitude of commercial scrapers and re-users like Google. Google incorporates material from Wikipedia on its own search results pages – in the Knowledge Graph panel – and is reportedly looking at increasing the amount of Wikipedia content it shows on its pages. This will maximise Google’s income from advertisements, as many users who are only searching for basic information will no longer see any need to leave the ad-rich google.com environment.
All of this content is given to Google and other re-users free of charge: by opting for a licence allowing commercial re-use, Wikipedia lost its chance to generate revenue from Google, Facebook and other commercial re-users of its content.
As we have seen, the Wikimedia Foundation has been doing perfectly fine financially anyway. The one party that does not earn anything from Wikipedia is the people who are writing it: the volunteers. Many still find that the work has its compensations – the satisfaction that comes from participation in a global free knowledge project, the hope to have given the world a well-researched and well-written article on an important topic. Writing and researching a Wikipedia article can expand one’s knowledge of a topic in a way few other online activities do.
There are less lofty compensations too: the pleasure of having increased the visibility of one’s own business or favourite band; promoting a political or ideological cause; editing for paying clients; adding a well-crafted hoax and seeing how long it lasts; or taking someone disliked down a notch or two by making their Wikipedia biography a little less flattering. It all goes into the Wikipedia mix.
But every Wikipedian incurs an opportunity cost when contributing to “the project”: it’s time spent on unpaid work, time spent away from friends and family, time spent not getting a career together. Many are socially isolated and hopelessly addicted to Wikipedia, spending many hours a day on the site.
Still wanting to donate?
If you are planning to donate to Wikipedia the next time you see a fundraising banner pop up, be very clear on what the effort is you are supporting. You are not supporting many thousands of rank-and-file volunteers creating and maintaining content you appreciate. Not a penny goes to them. (Perhaps someone should start a fund?)
Your money will not pay for any staff or outside experts tasked with checking and improving Wikipedia content. Out of the 215 employees of the Wikimedia Foundation, not a single one is given responsibility for evaluating the content of Wikipedia.
What you are supporting is in large part the further ambitious expansion of an organisation that does with paid staff what volunteers used to do for you for free, and which at present has alienated a significant part of the project’s volunteer community. And you’re gambling on the chance that the WMF’s new Executive Director will be able to turn the ship around, and ensure that your money is used in a way that delivers better value to you, the reader, than has been the case to date.
By all means buy a Wikimedia Foundation programmer a coffee this year. But be under no illusion that your money is used to keep Wikipedia online and ad-free. That is the one thing for which the WMF already has more than enough money.
Images: 1. Wikipedia (CC BY-SA 3.0); 2. Wikimedia Commons; 3. Wikipedia (CC BY-SA 3.0); 4. Wikimedia Commons, CC BY-SA-3.0; 5. By Don-kun, Wikimedia Commons, CC BY-SA; 6. Wikimedia Commons, CC BY-SA-3.0; 7. Wikimedia Commons, CC BY-SA 2.0.