[17:52:26] Hullo. [17:54:06] Welcome all. [17:54:16] We'll be getting started in a few minutes' time. [17:54:49] I've posted the stream link and the agenda in the /topic. [17:55:08] thank you forrester [17:55:08] As ever, if you have questions please ask them in here as soon as they occur to you. [17:55:39] Fellow viewers (even me) may be able to answer your question then and there, and if not, I'll ask it at the end (time permitting). [17:59:16] Moin [17:59:55] Plus [18:00:52] * Guest5667 waves to everyone. [18:01:04] o/ [18:01:26] Greetings everyone :) [18:01:44] _o/ [18:01:53] o/ [18:02:21] youtube stream did not start yet, right? [18:02:28] correct [18:03:08] thanks, there was a countdown, but it ended, but no stream, so /me was confused [18:03:33] yeah, zeljkof. I read that as delays. :) but I may be proven wrong. ;) [18:04:19] just about to get going here... [18:04:23] looks like youtube stream has started [18:04:41] but nothing is happening at the moment [18:04:45] Not yet. [18:04:53] Waiting for everyone to be ready. [18:04:58] Starting now. [18:05:07] noooo there's no sound! [18:05:13] sound now [18:05:15] I can see and hear the stream [18:05:26] Oh, there it is!± [18:05:32] I have some lag. Never mind! [18:06:36] Who's this "Varnum" guy [18:06:38] brendan_campbell: I hear some echo and Angel's voice is a bit interrupted. [18:07:05] brendan_campbell: echo is gone, and it's good now, I think. Thanks! [18:07:19] audio is there, but I lost video [18:07:21] foks: I've heard nothing but bad things :p [18:07:21] oh, it's back [18:08:08] okay, sorry for those hiccups [18:08:10] should be good now [18:08:17] np, brendan_campbell. all good. [18:08:21] brendan_campbell, champion! :D [18:08:31] I still have some echo I think, but it's not that bad [18:08:49] yeah, there is a bit of it Samwalton9. [18:11:20] Link should be https://wikimediafoundation.org/wiki/Non_discrimination_policy [18:13:19] James_F, question: How many people are joining overall? Would be good to see the gap as a percentage. [18:13:45] matt_flaschen: Thanks. Maybe someone here knows? Otherwise happy to ask. [18:16:32] Really cool project [18:16:44] do we have data on whether the facebook activities have a measurable impact on editors or readers? [18:17:06] gwicke: Will ask. [18:17:28] This is giving me so much warm fuzzies. [18:18:19] That would be a lot easier to track if we just had a Facebook like button on every page. *Way* better cross-site analytics [18:18:28] (Kidding obviously ;) ) [18:18:36] Tsk. :-) [18:18:59] matt_flaschen: Do these charts answer your question from before? [18:19:09] Wasn't there some news a year or two ago that promoted posts primarily brought in bot accounts? [18:19:15] * Samwalton9 goes to read... [18:20:05] gwicke: we have a research collaborator (Saiph Savage) who’s doing amazing work on social media engagement for prospective editors, focused on gender-related initiatives [18:20:09] James_F, technically, no, since I asked about the percentage of new people, and that's the overall percentage, but close enough. You can omit mine. There are probably better questions to ask. [18:20:10] see http://wvutoday-archive.wvu.edu/n/2016/01/19/wvu-researcher-creates-twitter-bot-platform-for-activists.html [18:20:17] Q: what do we know about effects of adding Facebook fans outside of Facebook itself (ie, converting them to new editors on-wiki, etc) [18:20:26] James_F, and it does show how bad the overall situation was, and that it's improving. [18:20:27] matt_flaschen: Kk. [18:20:31] she worked with WomenInRed and a bunch of equivalent initiatives in Spanish [18:20:50] oh, right, same Q as gwicke. carry on. [18:21:35] but would like to see an answer from the room, summarizing current understanding and/or theories for that Q. [18:21:37] gwicke, ragesoss, but in seriousness, it's rather hard to answer that question. I guess you would need surveys, unless people linked to their Facebook from their WP profile (which few would do) [18:21:49] Music in the background is great. [18:21:54] foks: +1 [18:21:56] matt_flaschen: yeah. I guess I want to hear about plans to measure it. [18:22:09] gwicke: http://www.saiph.org/lead.pdf [18:22:31] ^ this is a pilot of work in that area, which I advised [18:22:54] similar here; facebook likes are a means to an end, so I'm curious about the (eventual) impact [18:23:05] DarTar, cool, thanks for the info. [18:23:06] https://meta.wikimedia.org/wiki/Research:Guiding_and_Retaining_Volunteers_Using_Social_Media_Bots [18:23:12] I think this is what I was thinking of regarding the strange behaviour of likes coming from paid Facebook ads: http://www.bbc.co.uk/news/technology-18819338 [18:23:15] Thanks, DarTar. [18:23:16] matt_flaschen: because I think this is an area where, if we don't actually have plans around measurement of on-wiki results (whether readers or editors or whatever) or some other thing we care about outside of facebook... then the end result is that we're making Facebook a better experience for people, and paying Facebook for the privilege. [18:23:37] Also this: https://www.youtube.com/watch?v=oVfHeWTKjag [18:24:41] Like, I can think of plenty of effects, but Facebook engagement in and of itself isn't very interesting. [18:25:45] Not everything we do is trying to make people edit, of course. [18:25:56] ragesoss: I think the question we’re interested in if engagement in social media results in good quality contributions from newcomers and engagement in Wikimedia projects [18:26:02] Samwalton9, interesting. I can't think of any reason bots would click ads Facebook served to them, unless Facebook is running the bots. I haven't seen reports about that before, though. [18:26:20] foks: +1 as well [18:26:22] Samwalton9, that article is 5 years old though. [18:26:35] matt_flaschen, there's some facebook response here, saying that they've been working on it: https://www.washingtonpost.com/news/the-switch/wp/2014/02/10/this-blogger-paid-facebook-to-promote-his-page-he-got-80000-bogus-likes-instead/ [18:26:40] matt_flaschen: my understanding is that it's to shield their paid for clicks. Basically by clicking on a bunch of random ad's, they can click on 100k ads/like buttons/etc. that someone paid them to [18:26:41] I’m happy to say a few words about this project during the QA if there’s interest [18:26:59] DarTar: Thanks, I might call you on that. :-) [18:27:03] gwicke: ragesoss: +1 [18:27:06] kk [18:27:12] James_F: right. I mean, maybe it has general effects on populations in terms of how lots of folks understand Wikipedia. Or maybe it's pretty much just a silo. But I'm curious about what the bigger picture is in terms of why this is worth the large amount of effort it takes to do Facebook engagement well. [18:27:24] ebernhardson, ah, that makes more sense, thanks. [18:27:25] * James_F nods. [18:29:12] ragesoss: good questions. FWIW there was also https://blog.wikimedia.org/2012/06/20/social-media-for-wikipedia-outreach-india-program-pilot/ , but IIRC it was very high touch (a lot of work required to convert facebook users into productive editors) [18:30:20] a couple of examples of how recommender bots and wikiprojects are interacting on social media: [18:30:26] https://twitter.com/WikiWomenInRed/status/817895210902581249 [18:30:33] https://twitter.com/WikiWomenInRed/status/814767016380891137 [18:31:11] https://twitter.com/WikiWomenInRed/status/808039634705989635 [18:33:20] Lots of good directions here, too. [18:36:20] DarTar: happy to hear about social media experiments like the kinds you're pointing too, but also want to hear about plans/ambitions/theories/strategy for Facebook in particular. (Twitter engagement is a very different thing from Facebook engagement, in my experience.) [18:36:26] the facebook page has been useful to get in touch with engaged readers. people who love our brand and care about us, but maybe aren't showing up on IRC, wikimedia-l, village pumps, etc. [18:36:55] ragesoss: can’t speak to FB experiments, this pilot with Saiph mostly focused on Twitter [18:37:00] DarTar: has there been research about the impact of @wpstubs on edits? (i only know of the mention in this paper https://meta.wikimedia.org/wiki/Research:Newsletter/2015/July#.23Wikipedia_and_Twitter which said that while @wpstubs was the most active WIkipedia-linking bot on twitter, "the lack of edits suggests that the response of the community is limited") [18:37:38] Haeb: yes, I was going to mention that too. I should clarify that wpstubs is a hack, the work by Saiph is a controlled experiment :) [18:38:14] and i know of https://meta.wikimedia.org/wiki/Research:Newsletter/2016/October#cite_ref-14 [18:38:16] for the record that study predates the “adoption” of social media bots by wikiprojects such as WomenInRed [18:38:49] HaeB: yep [18:38:56] There's also ragesoss's awesome ORES-quality-based bot, though he's too modest to mention it. [18:39:33] that's completely untrue. I'm not too modest to mention. would not hestitate to toot own horn. [18:39:43] * James_F laughs. [18:39:44] it's been down for several weeks though. [18:39:46] :( [18:39:49] Yeah. :-( [18:40:25] wpstubs has been down for months too [18:40:38] https://twitter.com/WikiWomenInRed/status/840957189644775424 [18:41:12] The poor conversion on @wpstubs doesn't surprise me, to be honest. Unclear call to action and the level of effort needed to make a conversion (click --> ??? --> expanded article) is incredibly steep [18:41:16] * ragesoss reminds self to fix @Fixmebot soon. [18:41:20] (Right now I have gwicke and ragesoss's questions for Jeff/Aubrie around the impact of Facebook fans, and that's it – any questions for Marìa, Felix or Nicole?) [18:41:22] We're asking people to *write Wikipedia articles*, not click a like button or something. [18:42:01] hey harej: agree, like I said above, wpstubs is a hack :) [18:42:46] I’d love to hear ideas on different CTAs if they are cheap to implement [18:46:57] we have the referrer information, so we know who comes from Facebook. [18:47:19] Any other questions? [18:47:43] Even if there are ad-clicking bots, they're not getting money from WMF directly or indirectly. [18:48:06] Only Facebook is getting that money. [18:49:12] Asking a question right now is Chuck Roslof [18:49:25] Thank you, foks. [18:49:31] \o/ [18:49:34] If we're seeing the ads are resulting in improved engagement, then I guess it's not a concern. [18:50:35] DarTar: here's an idea for Twitter: a bot where I tweet it my Wikipedia page (draft, etc.) and tweets me back stuff about that page (a nice pageview graph, ORES stuff, AI suggestions for improvement, etc) [18:50:37] but are they Samwalton9? :) [18:50:38] Guest5667: wpstubs uses a campaign tag [18:50:40] are we* [18:50:47] a low tech method used by many sites to determine where new users come from is to give a subset of them a quick survey [18:51:10] Guest5667, I believe that's what Jeff just said, yes [18:51:30] I see. thanks, Samwalton9. [18:51:34] ragesoss, https://phabricator.wikimedia.org/T161236 [18:51:45] halfak: :D [18:51:54] :D [18:51:58] * halfak loves ideas not getting lost in the irc scroll :) [18:52:00] Any more questions? Last chance! [18:52:09] "Echo, help me improve this article” [18:52:19] that's step two, yeah. [18:52:27] or rather, step 3. [18:52:33] “Echo, improve this article” [18:52:33] step 2 is "..." [18:53:44] Thank you, Anna. :) And thanks to everyone who worked on this. [18:53:55] Policy, not just a draft now. :) [18:53:55] James_F, wikilove for Keilana and Rosiestep for their awesome work covering biographies about women on Wikipedia (highlighed recently in our blog https://blog.wikimedia.org/2017/03/07/the-keilana-effect/) [18:54:23] techconductcandidates@wikimedia.org (for Code of Conduct Committee candidates) [18:55:07] +1 juliet [18:56:33] DarTar, we need to make an Echo->Echo bot, where Amazon Echo reads your Echo notifications. [18:56:37] Mainly just to confuse people about whether it's confusing. [18:56:47] matt_flaschen: *cough* renaming [18:56:58] James_F, I keep emailing Amazon, but they don't answer. [18:57:12] :-P [18:57:37] Who did heatherw thank? [18:57:44] Oh! The "move" [18:57:45] the move folks. :) [18:57:46] halfak: Everyone involved in the office move. [18:57:46] matt_flaschen: as I understand it, the only thing that makes it up the chain is returning valuable merch. [18:57:47] Everyone involved in the [office] move [18:58:22] I guess return your "defective" Echo unit and maybe you'll get some traction? [19:03:52] "No Wikipedia editing support. Return to Sender"