FYI, I'm wondering if anyone has compared Wikipedia's hyper growth until 2007 and subsequent slower rate of production to the phenomenon of music CD sales in the 1980s/1990s​ that also surged then waned in a similar fashion. If you examine the graph of global music sales revenue since 1980, it will look familiar to those who study Wikipedia's new articles per year numbers. [1] [2] I noted this at a recent talk by Aram Sinnreich (Rutgers) who studies digital downloads and whether online music piracy killed CD sales. The conclusion was: no, piracy was not the major factor. The "format replacement cycle" of the 1980s/1990s was responsible for an unusual one-time surge, when people who owned vinyl or cassettes started re-buying titles they already had, to get high quality and convenient CD-versions. That "surge" was (erroneously) interpreted by the recording industry as the natural long term trajectory of sales, and the resulting music sales collapse has been falsely attributed to illegal downloads. Conclusion: the music sales curve is in fact the sum of several curves that take into account that one-time surge, how big box retailers priced their music, sales of $0.99 singles and the advent of new technologies like streaming services (Pandora, Spotify, etc). Using this model, Wikipedia experienced the same "low hanging fruit" surge in its early years. This is what Ed Chi (PARC/Google) considered the "catch up" phase of Wikipedia [3], when we were cataloging "common knowledge," and effectively replacing our "old" sources of info -- Britannica, Encarta, CIA Factbook, US Census, public domain, et al -- with content that was easy to input and energize new users. If we accept this as a plausible model vs the standard "hype cycle" [4] explanation of Wikipedia's rise/fall, we should look to other examples for which this "replacement" dynamic holds for more clues about Wikipedia's future. Phenomena we see in Wikipedia that affect community growth are remnants, and perhaps relics, of that earlier era. We need to ask: are policies adopted for 2007-era levels of hyper-participation that persist today harmful to ongoing community dynamics? (ie. Request for Adminship, restrictions on article creation, et al) In what instances should Wikipedia lower its shields, so to speak, in reaction to this new landscape? Do we as a community have the procedures or culture to be agile in response to these changing conditions? The tension between WMF R&D efforts and "the community" is a clear example of this inability to reconcile change (eg. visual editor and media viewer adoption). Not everything is gloomy, however. One could argue that the willingness of Europe-based Wikipedia editions to allow corporate representatives and "paid editors" to openly edit Wikipedia directly (an absolute no-no in en.wp) is an example of flexible policy when faced with the realities of long-term growth and maintainability. English Wikipedia certainly seems more rigid in that sense. That said, I still feel the "Facebook/Twitter ate my community" angle merits more analysis. The fact that all the major language Wikipedia editions dropped in that same 2007 time frame, as well as WikiHow, is still eyebrow-raising. It certainly demands we consider external effects. If you examine blogging activity during that same period, it dropped off a cliff from 2006-2009. From Pew (2010): "Only 14% of teens ages 12-17 worked on their own blog as of 2009, a drastic decrease since 2006, when twice as many (28%) said they had done so. Millennials have also seen a decline in blogging over the past couple years, from 20% in December 2008 to 18% in May 2010." [5] Those 12-17 year olds are now 17-22 years old, exactly the types of folks we'd like to get involved with Wikipedia. However, instead of being weaned on MySpace HTML hacking on a laptop screens like previous generation of "digital natives," they have grown up with slick WYSIWYG Facebook and Twitter interactions on their mobile device. Imagine what they think when they click "Edit" and see templates, info boxes and Wikimarkup galore. Don't stop at the "low hanging fruit" phenomenon. There are still lots of interesting dynamics to consider. PS: See what happens when your blogging activity drops off? You wind up spending an hour on an email to Wikiresearch-L [1] http://mcpress.media-commons.org/piracycrusade/chapter-5-bubbles-and-storms… [2] https://commons.wikimedia.org/wiki/File:Enwikipediagrowthcomparison.PNG [3] http://www-users.cs.umn.edu/~echi/papers/2009-WikiSym/wikipedia-slow-growth… [4] https://en.wikipedia.org/wiki/Hype_cycle [5] http://www.pewinternet.org/files/old-media/Files/Reports/2010/PIP_Generatio… _______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

wrote: I believe you are on something that would be very worthwhile to look further into I often think of three phases and likens it to shooting at a wall with a paintball gun -first you get enormous result and really see effect but the wall is uneven painted, but to fill in the gaps in not as fun (initial phase up to 2007) -then comes the filling in of empty spaces, take care of uneven paint etc. it needs more patience, precision and endurance (2007 up to about now) -But then I beleive we must enter a third phase, where we redo the painting but not with a paintball gun but with a painting machine, that meticulous and methodology repaint the wall board by board, where coverage, sustainable and quality is at focus. And here I see bots, Wikidata and semiautomatic efforts as key components And I would love to get the first two of these phases validated by numbers as you think of. And later then reflect on how we work as communities, also having in mind the third phase I anticipate as necessary Anders Andrew Lih skrev den 2014-12-12 12:40: FYI, I'm wondering if anyone has compared Wikipedia's hyper growth until 2007 and subsequent slower rate of production to the phenomenon of music CD sales in the 1980s/1990s​ that also surged then waned in a similar fashion. If you examine the graph of global music sales revenue since 1980, it will look familiar to those who study Wikipedia's new articles per year numbers. [1] [2] I noted this at a recent talk by Aram Sinnreich (Rutgers) who studies digital downloads and whether online music piracy killed CD sales. The conclusion was: no, piracy was not the major factor. The "format replacement cycle" of the 1980s/1990s was responsible for an unusual one-time surge, when people who owned vinyl or cassettes started re-buying titles they already had, to get high quality and convenient CD-versions. That "surge" was (erroneously) interpreted by the recording industry as the natural long term trajectory of sales, and the resulting music sales collapse has been falsely attributed to illegal downloads. Conclusion: the music sales curve is in fact the sum of several curves that take into account that one-time surge, how big box retailers priced their music, sales of $0.99 singles and the advent of new technologies like streaming services (Pandora, Spotify, etc). Using this model, Wikipedia experienced the same "low hanging fruit" surge in its early years. This is what Ed Chi (PARC/Google) considered the "catch up" phase of Wikipedia [3], when we were cataloging "common knowledge," and effectively replacing our "old" sources of info -- Britannica, Encarta, CIA Factbook, US Census, public domain, et al -- with content that was easy to input and energize new users. If we accept this as a plausible model vs the standard "hype cycle" [4] explanation of Wikipedia's rise/fall, we should look to other examples for which this "replacement" dynamic holds for more clues about Wikipedia's future. Phenomena we see in Wikipedia that affect community growth are remnants, and perhaps relics, of that earlier era. We need to ask: are policies adopted for 2007-era levels of hyper-participation that persist today harmful to ongoing community dynamics? (ie. Request for Adminship, restrictions on article creation, et al) In what instances should Wikipedia lower its shields, so to speak, in reaction to this new landscape? Do we as a community have the procedures or culture to be agile in response to these changing conditions? The tension between WMF R&D efforts and "the community" is a clear example of this inability to reconcile change (eg. visual editor and media viewer adoption). Not everything is gloomy, however. One could argue that the willingness of Europe-based Wikipedia editions to allow corporate representatives and "paid editors" to openly edit Wikipedia directly (an absolute no-no in en.wp) is an example of flexible policy when faced with the realities of long-term growth and maintainability. English Wikipedia certainly seems more rigid in that sense. That said, I still feel the "Facebook/Twitter ate my community" angle merits more analysis. The fact that all the major language Wikipedia editions dropped in that same 2007 time frame, as well as WikiHow, is still eyebrow-raising. It certainly demands we consider external effects. If you examine blogging activity during that same period, it dropped off a cliff from 2006-2009. From Pew (2010): "Only 14% of teens ages 12-17 worked on their own blog as of 2009, a drastic decrease since 2006, when twice as many (28%) said they had done so. Millennials have also seen a decline in blogging over the past couple years, from 20% in December 2008 to 18% in May 2010." [5] Those 12-17 year olds are now 17-22 years old, exactly the types of folks we'd like to get involved with Wikipedia. However, instead of being weaned on MySpace HTML hacking on a laptop screens like previous generation of "digital natives," they have grown up with slick WYSIWYG Facebook and Twitter interactions on their mobile device. Imagine what they think when they click "Edit" and see templates, info boxes and Wikimarkup galore. Don't stop at the "low hanging fruit" phenomenon. There are still lots of interesting dynamics to consider. PS: See what happens when your blogging activity drops off? You wind up spending an hour on an email to Wikiresearch-L [1] http://mcpress.media-commons.org/piracycrusade/chapter-5-bubbles-and-storms… [2] https://commons.wikimedia.org/wiki/File:Enwikipediagrowthcomparison.PNG [3] http://www-users.cs.umn.edu/~echi/papers/2009-WikiSym/wikipedia-slow-growth… [4] https://en.wikipedia.org/wiki/Hype_cycle [5] http://www.pewinternet.org/files/old-media/Files/Reports/2010/PIP_Generatio… _______________________________________________ Wiki-research-l mailing listWiki-research-l@lists.wikimedia.orghttps://lists.wikimedia.org/mailman/listinfo/wiki-research-l _______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l

On Fri, Dec 12, 2014 at 3:40 AM, Andrew Lih <andrew.lih(a)gmail.com> wrote: Andrew - Do you have a source for the wikiHow data? I've been with the project as part of the staff since it started and the community manager since 2010 and don't recall a 2007 drop. wikiHow started in 2005 and I think the community was still seeing growth at that time, though we only started tracking metrics in 2009. Our 5+ edits trend since then looks like this: [image: Inline image 1] Please don't republish this graph - the data here is rather shaky (issues like how [in]consistently it included user page edits or paid translator edits, there is room for a project here gathering more accurate historical data to compare apples to apples). It just gives a rough idea that's pretty consistent with my anecdotal impressions from having been involved with the community since it began. It's worth noting that in 2010 we started launching some on-ramping tools that made it much easier to get to the fifth edit. I.e. you could easily get to 5 edits with 5 clicks. _______________________________________________ Wiki-research-l mailing list Wiki-research-l(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l