The Signpost

News and notes

A future with a for-profit subsidiary?

Contribute  —  
Share this
By Smallbones and WereSpielChequers

For-profit subsidiary started by WMF

The non-profit Wikimedia Foundation has incorporated a for-profit subsidiary, Wikimedia LLC in order to make money from Big Tech companies like Alphabet Inc.'s Google and YouTube subsidiaries, as well as Amazon, Apple, Microsoft and Facebook. These companies are intensive users of Wikipedia content for products such as the Google Knowledge Graph, Amazon Alexa, and Apple's Siri. Wikimedia LLC plans to offer a service called Enterprise API to provide these and similar companies with speedy tailored access to Wikipedia data. Wikipedians and the WMF have long called for Big Tech to donate more money to the WMF to support the production of the freely licensed content they use. Some Big Tech companies do make occasional donations, but the total appears to be less than $5 million per year.

Big Tech and other paying customers would benefit by having more business-like service-level agreements (SLAs), getting clean timely data that meets their specific needs. Other users would benefit from having full, but less timely, access to the same data that Big Tech uses. The WMF would benefit by getting more, and more consistent, revenues that they can use to fund the aggressive growth plans described in the strategic plan. The overall Wikipedia movement would benefit by having broader and more timely distribution of our freely licensed content, with our content licensing requirements respected.

It's common in the U.S. for non-profit organizations to own for-profit subsidiaries. For example, non-profit museums may own book stores and restaurants designed to support their cultural mission and raise cash. Almost any type of for-profit business can be owned as long as the business supports the overall mission of the non-profit, even if that is only by raising cash. For example the non-profit Hershey Trust Company, a charity that helps educate disadvantaged children, effectively controls the Hershey Company, a multinational food and chocolate company, and owns Hersheypark and several other tourism related companies. Taxes must be paid on the for-profits' earnings.

The WMF has discussed the project with Big Tech to determine that they have potential customers and the general areas of their interest, but no agreements have been reached. The WMF is also consulting Wikipedians on the project. Currently the most convenient place to ask questions and give feedback is on the meta talk page for Wikimedia Enterprise. More structured consultations will be scheduled. WMF spokesperson Liam Wyatt says the timing of the consultations was set in order to ensure that enough solid information is available to allow meaningful discussion, while still having the input be useful for determining the overall shape and many details of the proposed project.

Wikipedians' reactions to the announcement, as usual, are quite varied. Many Wikipedians are happy that Big Tech will be paying their fair share of WMF's expenses. Administrator Johnuniq states "An enterprise API as a fee-for-service is desirable and long-term funding would be good". Others seem surprised that the WMF can take on such a project. From legal, practical, and historical perspectives, it can and already has taken on a similar project. Strategic, ethical, and financial points of view may lead to more productive discussions. The principles that the WMF has laid out for the project are described on meta.

The earlier project was described to The Signpost by the WMF's first employee, Brion Vibber. Leaving out the technical details, Vibber wrote "The start year was 2004. It didn't scale super well because you either had to parse the wikitext yourself or hook it up to a MediaWiki instance and all that entails... Eventually the potential clients found their own workflows... And then eventually they came back and decided they wanted us to make a better workflow, and the Enterprise API was born!"

Long-time Wikipedian DGG notes that Wikipedia and Google have had a long and mutually beneficial relationship.

We would be less reachable by the public without the traffic that comes from them, but they would be much less valuable without having at least some fairly decent content, a great deal of which comes from us. Possibly this degree of relatedness is worth preserving: If we look at ultimate value to the general public, having our content on Google improves the amount of somewhat reliable information available there, especially on topics they wouldn't otherwise have any content-- in that sense, they act as a co-distributor. As an illustration, Google content derived from WP might be available in countries that prohibit access directly to WP.

A 2005 Signpost story illustrates the sometimes close relationship well. Wikipedia was having trouble keeping up with the load on its servers and Google offered to host some of Wikipedia for free on its servers. As Jimmy Wales recalled to The Signpost he was a guest lecturer at Stanford and ran into Google founders Sergey Brin and Larry Page. They offered the use of Google’s servers but there were unforeseen technical difficulties, so they donated money instead to the WMF.

Still DGG is very skeptical about the WMF's need to raise more money and fears "commercial entanglement". User Guy Macon has similar views, believing that raising more money only distracts from the needed focus on the volunteer community. When asked if the WMF could do anything to earn his support, he answered.

Do it without any money from donors or employee time paid for by donors. Every other for-profit LLC has to secure funding somehow; why should this one be different? Make a pitch to Google, Amazon, and Apple. If they are willing to pay for the service in the future they should be willing to invest in completely funding the creation of the service now.

Lockdown continues – uptick in Wikipedia editing

As the world comes to 12 months of lockdowns, we recall sofa surfing, socially distanced dating and for many the great declutter. Wikipedia editing has joined the ranks of indoor speleology, gardening, baking and mask making as one of those activities that people have resorted to because they still can. Early signs are that 2021 has started with an editing frequency that the community hasn't seen since 2010 (apart from one spike when we moved the interwiki links to Wikidata). W



Do you want to contribute to "In the media" by writing a story or even just an "in brief" item? Edit next week's edition in the Newsroom or leave a tip on the suggestions page.
S
In this issue
+ Add a comment

Discuss this story

Wikimedia LLC/Enterprise API

[1] editors to pee in a bottle. -DePiep (talk) 21:38, 28 March 2021 (UTC)[reply]
It looks like meta: Talk:Wikimedia Enterprise#Letting the wolves in at the doorBri (talk) 23:41, 28 March 2021 (UTC)[reply]

Uptick in editing

This is newsworthy. But I'm not sure why the Signpost should rely on the rather indirect "Time between edits" metric which has a lot of shortcomings (e.g. unable to exclude bot edits or to focus on mainspace edits), when filtered monthly edit count stats are readily available now and show the same thing more clearly.

(By the way, it should be noted that the while the volume of edits has been increasing, the number of active editors remains stagnant. Still, its former downwards trend that gave rise to so much "decline of Wikipedia" discussion and theorizing has already stopped about half a decade ago.)

Regards, HaeB (talk) 21:26, 28 March 2021 (UTC)[reply]

Interesting. Seems to line up with the increase in users making 100+ edits per month since the start of the pandemic. Perhaps a reflection of increased time for superusers (or use of WP editing as a stress coping mechanism). The more casual user brackets (1-4 and 5-24 edits per month) remain stubbornly flat. T.Shafee(Evo&Evo)talk 23:23, 28 March 2021 (UTC)[reply]
But how much is just administration? Categories, short description, tagging, general fixes (See Also --> See also) and so on. It could be nice with some statistics... Christian75 (talk) 23:42, 28 March 2021 (UTC)[reply]
I like Time Between Edits because it is such a Direct measure. You know it isn't going to be effected by things such as the creation of draftspace. True at least two spikes in it over the years have been attributed to bot activity, but I find it interesting to log that as well. And yes it isn't the only stat - if anyone had time available to expand the story then looking at another stat would have been a logical route. ϢereSpielChequers 08:49, 29 March 2021 (UTC)[reply]


@WereSpielChequers: Can you please take another look at the first sentence of this single-paragraph section in the singpost? It appears to be incomplete and I'm not sure what you're trying to say so I can't make any suggestions to fix it. Specifically, it seems to be lacking a verb. Thanks! ElKevbo (talk) 01:12, 29 March 2021 (UTC)[reply]

I noticed that pre-publication but let it slide. Perhaps As the world comes to 12 months of lockdowns, we recall ...Bri (talk) 01:16, 29 March 2021 (UTC)[reply]
That works for me. ϢereSpielChequers 08:49, 29 March 2021 (UTC)[reply]
I wondered if the 2021 uptick might be linked to the "race for 6M" articles on English wikipedia. That was clearly an internal "campaign" which would account for the regular contributors making higher contributions than usual rather than attracting new contributors. We've had COVID lockdowns in my country on and off since March 2020. If lockdowns were the cause of increased editing, the uptick would have started back in 2020. I don't think an increase in 2021 can be attributed solely to lockdowns. Kerry (talk) 03:27, 29 March 2021 (UTC)[reply]
Our 6 millionth article came last January at a time when we were doing 10 million edits every 62 days or so, so I don't see a link. The uptick did start a year ago, and I put that in the Signpost, but this latest spike went above anything we saw last year. ϢereSpielChequers 08:49, 29 March 2021 (UTC)[reply]
Article creation is a drop in the ocean compared to the many edits that the grinders make. They were chasing the billionth edit which was this. Andrew🐉(talk) 18:19, 29 March 2021 (UTC)[reply]
By the way, a new preprint that came out a week ago examines exactly this question in a more thorough fashion (finding that "contributions to the English Wikipedia increased by over 20% compared to the expectation derived from pre-pandemic data"). If anyone is interested in reviewing it for next month's "Recent research", feel free to sign up for it here (search for "Ruprechter"). Regards, HaeB (talk) 09:14, 29 March 2021 (UTC)[reply]





       

The Signpost · written by many · served by Sinepost V0.9 · 🄯 CC-BY-SA 4.0