[17:47:18] Hey everyone. Metrics will being in ~13 minutes' time. [17:49:39] I'll be IRC functionary as is traditional. If you have a question you want answered shout it out in here. I'll try to get it answered, either from the wonderful fellow IRCers in here or via Q&A at the end of the meeting. [17:55:50] hi [17:56:27] hi rfarrand [17:57:56] :) [17:58:51] actually is or was the general secretary of the chapter [17:59:46] (sorry, wrong channel) [18:00:26] Some slight technical difficulties, sorry. [18:00:56] Metrics starts late, news at 11 [18:01:06] marktraceur: Be nice. ;-) [18:02:33] OK, we're live. Welcome everyone. [18:02:57] Jake is our host. [18:03:08] In his fabulous video call room. [18:03:37] Indeed. [18:03:48] Whoa, I didn't realize pajamas were an option. [18:04:10] This should be fun :-) [18:05:17] Congralations to Katy Perry! [18:05:18] ha ha ... [18:05:24] :P [18:05:26] Whoops. [18:10:15] in the youtube chat area, what does "Create a channel to join the chat" mean? (i'm logged into my google account already but this msg appears when hovering over "say something") [18:10:33] HaeB: Maybe you haven't created a YouTube account yet? [18:10:39] (also, not enthusiastic about us continuing to split communication channels between here and YT) [18:10:50] yeah, HaeB. I had the same question yesterday in research showcase. I couldn't type. [18:10:52] nice summary Daniel [18:11:02] full report and a blog post on WikiCite to follow [18:11:13] Thanks, DarTar! [18:11:32] i see my avatar in the top right corner, so i thought i'm logged into my youtube account already [18:11:38] we have official laptop stickers too, come and find me if you want one [18:11:58] HaeB: Google accounts and YouTube accounts have a complex relationship. [18:12:18] HaeB, there's a separate layer of YouTube account. You'll have to follow the instructions if you want to comment: https://support.google.com/youtube/answer/1646861?hl=en [18:12:33] And they insist on calling accounts 'channels'. [18:12:35] HaeB: Given our issue with SULF I'm minded to give them some leeway, but… [18:12:39] Yes, what matt_flaschen said. [18:12:42] I don't see Ilya's screen. [18:12:50] is it okay in YouTube? [18:12:54] me neither [18:12:55] nope [18:12:59] I see it now... [18:13:00] brendan_campbell: ^ [18:13:05] yes, it is okay now. [18:13:06] can some one else share it? [18:13:20] I see his slides via the stream (and locally). [18:13:22] ok, so brendan_campbell fixed it for YouTube. In Hangout I still don't see it. [18:13:37] can someone please share their deck in the hangout? [18:13:39] i see it now [18:13:43] too many things for me to juggle [18:15:35] i see, thanks matt_flaschen. (i guess that's fallout from the partial abandonment of google plus) [18:15:38] I did brendan_campbell [18:15:49] Thanks, leila. [18:15:49] thanks leila [18:16:04] Can someone add these links to the agenda? https://meta.wikimedia.org/wiki/Wikimedia_Foundation_metrics_and_activities_meetings [18:16:27] matt_flaschen: The link from the slide? [18:17:00] James_F, there were three links a slide or two ago, including a guide for setting up hackathons. [18:17:12] Oh, right. They'll be in the slides, right? [18:18:20] James_F, yeah, I guess I'll come back to it. [18:18:51] Oh, this is really interesting (the high-awareness brand) [18:19:27] Anecdotally, from my experience, a surprisingly large fraction of people have no idea how it works. Great to see this study. [18:21:29] What comes after Generation Z? Generation AA? [18:21:45] * funcrunch is a Gen X'r, I think [18:22:19] funcrunch: Gen א I would say. [18:23:03] I used to be Generation Y, now I'm a Millenial. Not sure how the name 'Millenial' makes any sense. [18:23:31] is there a youtube stream? [18:23:42] But using letters for generations is like naming something post-something. The poor planning makes me question their whole thesis. [18:23:46] Superyetkin: Yes, it's in the topic: https://www.youtube.com/watch?v=mG6fBk78VRQ [18:24:15] oh, the video cannot be played on Firefox? [18:24:23] And how many of that 40% (85% - 45%) actually have Wikipedia in top 3, but access it via Google? [18:24:34] I'm watching it on Firefox @Superyetkin [18:24:42] Me too. [18:24:46] any codec needed? [18:24:48] Superyetkin, try youtube.com/html5 . [18:25:05] matt_flaschen: If 1960s is X, then A = 1960-23*25 = 1385s? [18:25:10] 25 to 30% look on facebook for *information*? scary... [18:25:17] I'm surprised *as many* as 50% know we're a nonprofit. [18:25:30] harej: I suppose the banners have had their effect. [18:25:34] Yeah, I think that's an improvement. [18:25:50] Can't prove it, though. :) [18:26:05] matt_flaschen: apparently the name relates to having a high school graduating class > 2000 [18:26:28] Gen Z? [18:26:31] Or Gen Y? [18:26:35] How much to folks know about other non-profits they work with or utilize? [18:26:36] first generation to have a graduating class in the new millennium [18:26:38] That's way earlier than I would have guessed. Math is hard. :) [18:26:42] Gen Y [18:26:46] OIC [18:26:55] it's like windows naming... X, Y, Millenium... [18:27:05] https://en.wikipedia.org/wiki/Millennials#Terminology [18:27:12] Oh, there are actual age ranges on this slide. :) [18:27:34] Didn't they say because it's not a scientific sample we can only use it for directional data? [18:27:41] Yes. [18:28:03] James_F, can you ask whether the study's science allows drawing these fine distinctions (e.g. 8.0 vs. 8.3)? [18:28:14] I will. [18:28:42] https://en.wikipedia.org/wiki/Generation_Z: "Generation Z (also known as iGeneration, Post-Millennials, Plurals, or the Homeland Generation in the United States)" [18:28:47] "Plurals"?? [18:29:00] CKoerner_WMF: Good question. [18:29:00] what on earth is the homeland generation? [18:29:04] are you using any extensions on Firefox? [18:29:06] ^^ [18:29:11] They should add "I'm not an an expert in anything" as an option. That's a common given reason in my experience. [18:29:31] krmaher: as a child i remember the point that "homeland" entered the national discourse [18:29:38] krmaher: Isn't that the ghastly term for 2001-terrorism-affected-young-adults? [18:29:44] whereas before I suppose it wasn't really a thing [18:30:04] James_F that's an utterly awful way of pigeonholing a generation. [18:30:15] we're used to it [18:30:23] krmaher: I know. I find it personally not just awful but actively offensive. [18:30:28] millennials: the cause of, and the solution to, everyone's problems [18:30:35] :> [18:31:00] krmaher, it is pretty awful. Department of Homeland Security has existed most/all (depending birth year) their aware life (assuming people have minimal memories before 6). Also, "we have always been at war with East Central Asia" [18:31:01] According to the article there was "online contest in which respondents voted overwhelmingly for the name Homeland Generation." [18:31:04] harej, that's true, it wasn't a thing for my generation. /me she says like a truly wizened elder [18:31:06] Sad to think about. [18:31:44] then again, I'm apparently an Xennial, which is also just... [18:31:57] krmaher: i had an excellent conversation with urvi about how our generation is defined by childhood trauma. fun!!!! [18:32:00] Yahoo and Yandex are not suprising if you know those markets [18:32:06] you're Ubuntu Xenial? [18:32:09] they are the dominant search providers [18:32:21] Let's go with "Plurals" instead even though that's also a stupid name; it's a less depressing one. [18:32:26] Is that from Scientology? [18:32:26] most of the value of Yahoo was Yahoo Japan+alibaba [18:32:53] harej :( [18:33:48] Good morning. When we get to Q&A, I have a question related to the readership research. Is someone available to queue questions for Q&A? [18:34:05] Pine: Yes, still me. [18:35:07] Ah, great. My question is: What are the primary objectives of the readership research? Are the objectives to increase fundraising revenue, improve dissemination of information, or some combination of those, or some additional objectives? [18:35:23] Here is the study Pine: https://cs.stanford.edu/people/jure/pubs/wikipedia-www17.pdf [18:36:09] Pine: full description is on Meta:Research [18:36:10] Thanks Sadads. [18:36:15] with goals, next directions etc [18:36:20] Cool. [18:36:25] Thanks Sadads and DarTar. [18:36:27] let me look it up for you [18:36:33] WWW paper is just one output [18:37:28] Pine: https://meta.wikimedia.org/wiki/Research:Characterizing_Wikipedia_Reader_Behaviour [18:37:53] “The outcomes of this research can help the community at large, Wikipedia's editor and developer communities, as well as the Wikimedia Foundation (e.g., the Reading team), to make more informed decisions about how to create and serve encyclopedic content in ways that are more suitable for the needs of those who seek to access it, and to design appropriate user experiences.” [18:38:37] we’ve basically built an online encyclopedia without knowing for years how people read it and navigate it, whether they find what they are looking for, etc [18:38:50] that’s what this work is mostly meant to inform [18:39:35] I see. It seems that fundraising is very interested in readership research too, which understandable, and I don't want to sound like I'm complaining about Fundraising because I recognize that it's necessary, but it's good to get a big-picture understanding of what the goals are here. [18:40:47] It crossed my mind (and maybe James_F could think about how to word this maybe for a question for a presenters) to wonder if it would be beneficial to differentiate the Wikimedia reader interface for different audiences based on geography, use case, etc. [18:41:03] I suppose that could be a topic for further research. [18:41:30] Pine: If you want to propose research projects for next year's annual plan, this is not really the venue. :-) [18:41:40] Hah, well, I'm not that far ahead yet. [18:42:09] :D [18:42:25] I might queue that question up, though, under some list of possible questions for future research at some indeterminate time. [18:44:02] Pine: I will say we're thinking about interfaces for different experiences (mobile and apps) that would help better serve these motivations and needs. Its unlikely that will be geo-specific interfaces, but rather task/motivation informed improvements and features [18:44:02] (PS to the above about YouTube: turns out that there is also a direct link to creating a channel, not given in the help page: https://www.youtube.com/create_channel ) [18:45:07] JoshM: cool, thanks. Those of us who are contributors have frequent discussions about editing interfaces for different use cases, and the same thought process could probably be applied to readers. [18:45:07] (speaking for the Readers team ^ which is one of the main motivators/consumers of this research) [18:45:37] Would it be practical to infer motivation for users without surveying them (based on the patterns of different use that Leila discussed), and adjust the user experience based on which type of user they are inferred to be? [18:46:09] Like, the 'related articles' that would be most enticing might differ between the bored user and the here-to-learn user. [18:47:25] ragesoss: Might one human be multiple user types at different times? Bored vs. school work, for instance. Might be hard to generalise? [18:47:28] James_F: I imagine you’ll relay these Qs to leila, so I’ll let her respond [18:47:29] Wikiclippy: "It looks like you're studying for mid-terms. Would you like help with that?" :D [18:47:41] lol [18:47:54] CKoerner_WMF: patented already [18:48:11] Pine: when you "What are the primary objectives of the readership research?" which research are you referring to? basically, say if you want me to answer or John. :) [18:48:20] DarTar: Yes, though Leila's here and most of them are mostly answered already, it seems. Might de-prioritise. [18:48:32] James_F: yeah, I just ask because it brings interesting possibilties to mind. The story of different motivations tracking with significantly different ways of using Wikipedia brings up a lot of possibilies. [18:48:50] James_F, you can detect sessions and do it by session. [18:48:53] leila: either is fine, thanks. [18:48:55] It would be interesting, though potentially confusing. [18:48:56] yo lz, nice preso [18:49:02] thats the direction we'd like to head for "power readers", but not quite to that level of sophistication yet. we also have to be very careful to maintain privacy and user-control in any approach, which makes this more complex for us than [18:49:19] other systems which are more personalized [18:49:32] so hard to monitor two discussions here and on YouTube, while also watching the stream [18:49:34] matt_flaschen: Yeah, but if you can tell only after e.g. the third page it's too late to retrospectively show different content styles. Interesting challenge for JoshM et al. [18:49:36] Encyclopedia isn't written by the same people as wikimediafoundation.org [18:50:05] +1 to anything that helps us collaborate better with ally organizations and share our broader views (free knowledge, open standards, anti-censorship, etc.) [18:51:23] Documentation "often" impedes collaboration? Wat? [18:51:27] I like that our contractor is essentially advocating https://en.wikipedia.org/wiki/Wikipedia:Ignore_all_rules for our official corporate site. [18:51:44] :) [18:51:45] that is interesting about documentation impeding collabration ...and a bit baffling. [18:51:54] hahaha, indeed @matt_flaschen [18:52:06] In seriousness, I would like this explained a little better though. [18:52:12] me too. :) [18:52:22] DarTar: agree that presentation + side discussions are difficult to follow simultaneously. I'll probably re-watch portions of the presentations that I've missed. [18:52:32] oh, was matt_flaschen referring to the talk or the wiki page? [18:52:37] @subbu its a pretty well established principle: because documentation is about reducing the amount of time that you have to spend repeating interactions with others. [18:52:42] I get the point about rules & jargon creep, but… I mean, that's not the right way to convey this. [18:52:49] subbu? I was referring to "Documentation impedes Collaboration" [18:52:55] thanks, DarTar (I'm catching up with IRC:) [18:52:57] Sadads: You're doing documentation wrong :) [18:53:27] ah, the rtfm effect? [18:53:35] I really don't get "collaboration [is] no substitute for communication" [18:53:52] Honestly, I don't completely understand the thesis, but that's partly because I'm multitasking between the chat and fully listening. [18:53:58] @Subbu yeah, that is what I would suggest. [18:54:21] My general idea was that they're saying people are too slow since they follow overly-detailed tedious rules. [18:54:30] matt_flaschen and DarTar: want to join me for a re-viewing of the presentation later? [18:54:42] Plus points for crediting the photo and the screenshot! [18:54:48] (Although we'll probably be chatting then, too, so....) [18:54:55] James_F: I have a question for John. Where do we want the awareness of our brands be? I can imagine that we want 100% of respondents to know wikipedia is written by volunteers. but do we also want the same ratio know the Foundation exists? or it's non-profit? Given that the Foundation has a supporting role, I can imagine more emphasis can be given to the projects' awareness versus the Foundation [18:54:59] Yeah... I'm going to rewatch the first 8 minutes, though, so we can do that. [18:55:06] matt_flaschen: Short version that might be wrong from me: "Too much generalised content overwhelms without helping people do specific tasks. We should instead provide clear, short answers to the direct questions they have." [18:55:15] leila: Thanks, in the list. [18:55:16] Pine: I may not be following the discussion here later this afternoon [18:55:22] thanks, James_F. [18:55:28] * subbu steps away for a meeting in 5 mins .. and refocus [18:55:37] Agree, though, in my experience we have "clear, overly long answers to every possible specific topic" (at least on WP, not sure about foundation wiki) [18:55:45] We're also very over time already. :-( [18:55:55] great job, leila. really interesting stuff. [18:56:11] James_F: that's roughly the answer I arrived at intuitively, too. The Wikipedia Way is "read a thousand pages of documentation" and a lot less human, personal, interactive. [18:56:17] Yup. [18:56:19] thanks ragesoss. I will answer your question, here or there. Let's see what happens with questions. [18:56:38] James_F: brendan_campbell: given that this is all remote can we go over time? :D [18:56:39] leila, I would like people to know it's 100% written by volunteers and what exactly donors are actually spending their money on. Hard to explain, buyt important. [18:57:02] leila: That's not how it works. Being remote means it's more likely people have clashing events immediately afterwards. [18:57:03] James_F: not a problem with me...but idt i'm the one to make that call [18:57:13] And community members too. The Foundation does a lot of work that isn't always communicated well. [18:57:19] matt_flaschen: agreed, though the second part of your sentence I'm with if you don't want to go to the penny. ;) [18:57:27] I can go a few minutes over, but many of our audience might not be able so to do. [18:57:35] :) [18:57:53] * leila is sad that James_F says this is not how it works. :/ but yeah, I get it. [18:57:57] * Pine will be re-watching portions of the presentations afterwards anyway and so will be hanging around this channel for awhile [18:58:25] * Pine ...unless something blows up, which does happen from time to time. [18:58:29] Actually, I have a meeting too, now that you mention it. [18:58:49] Question about the Mule research project: Did this also involve examining the current readership data for wikimediafoundation.org (i.e. which parts of the site get how many pageviews, are most popular etc)? [18:58:52] I like "quirky" [18:58:56] that's us [18:59:04] vcoleman: :) [18:59:39] In some ways we're quirky, in other ways very much buttoned-up. A little of both is good. I wouldn't really want our public corporate presence to be quirky, though. [19:00:27] HaeB: Do you have that data that you could share with us? [19:00:36] matt_flaschen: there is definitely diversity in formality. I would refer to Wikimedia DC as "Wikipedia in dress shirts." There are other aspects of the movement I would definitely not characterize as that. [19:00:39] http://tools.wmflabs.org/topviews/?project=wikimediafoundation.org&platform=all-access&date=last-month&excludes= [19:00:42] can we at least have a small quirky box? [19:00:57] We should be a little stuffy. Something that hints "recreational lawyer". [19:01:06] Thanks for the question asking and answers [19:01:16] and what about all our dogs? I would like them featured in our quirky box! [19:01:42] cf. https://www.amazon.com/p/feature/d8j6je99gnx65bk [19:01:46] WikiCite seems fascinating. How will it be used in practice? E.g., will we one day be able to compare an article about a book to the WikiCite model and determine whether the article contains all the data that the model says it should? Or if not that, what is the idea? [19:01:59] HaeB: I guess your q. is answered ? [19:02:29] so the answer is no? [19:02:32] yes! even our quirky is corporate. Sigh. [19:02:35] vcoleman: Like this? http://www.sanger.dk/ [19:03:09] We could put that in a box. [19:03:20] it's a question about the scope of the project, so it would still be useful to know mule's take on it [19:03:41] heatherw: there is data, yes [19:03:46] K4-713: WMF.org hompage design is now done :) [19:03:51] thanks for asking, James_F. Juliet and John: thanks for responses. [19:04:30] Any last questions? [19:04:46] K4-713 vcoleman we could use abittaker's kitten cam [19:05:08] YES!!! [19:05:25] Nice. :) [19:05:26] Oh dear. We're sum-of-all-knowledge-plus-KITTENS! now? ;-) [19:06:17] https://commons.wikimedia.org/wiki/Category:Felis_silvestris_catus [19:06:28] Any WikiLove people want me to convey? [19:06:40] ragesoss: If I understood your question correctly, it will boil down to: can we predict user motivations using the data we have? is this correct? [19:06:46] Kitten knowledge is implicitly included in all knowledge :) [19:06:54] brion: Yeah yeah. [19:06:56] leila: yes, good summary. [19:07:03] Yes, to Lynette and the move team [19:07:10] WikiLove for Ellie and everyone working on Wikimania [19:07:20] James_F: my Wikilove would be that I'm glad to see the work happening with structured data on Commons, and on the recent changes feed. :) [19:07:35] WikiLove for Robert and Lynette and everyone working on the move [19:07:37] Pine: We live to serve. :-) [19:08:28] More wikilove to David Strine for bringing Jasper to the office [19:08:30] Gonna pass on WikiLove to the elections committee and candidates who gave me feedback on the cycle this year. :) [19:08:39] WikiLove for Caroline Sinders - check out her recent writing [19:08:59] thank you all! [19:09:15] heatherw: recall https://phabricator.wikimedia.org/T51266#1563210 - that data is still there [19:09:16] leila: and relatedly (whether the answer is yes or no), do you think these are distinct enough personas to have a life beyond just the research project? Like, having different types of readers to systematically be talking about and keeping in mind could potentially be very useful in figuring out how to make Wikipedia better, both in terms of content development and product. [19:09:52] ragesoss: got it. in this case, here is my response. we tested with the data to see if the features we have are predictive of the three dimensions and their categories. The answer was that we can predict, but the predictions are too poor to be useful (in the best cases we would get around 60% AUC). We spent maybe a few weeks on this, so it's limited research, but one issue seems to be that as you say: people change motivations ac [19:10:44] ragesoss: We can spend resources and try to get to the bottom of it, but we 'd like to have a reason for doing this (for example if we know a product can heavily benefit from such research). otherwise, it will be a nice research problem, but the knowledge generation from it can be too limited. [19:11:51] heatherw: You can see pageviews in this tool: http://tools.wmflabs.org/siteviews/?platform=all-access&source=pageviews&agent=user&range=latest-20&sites=wikimediafoundation.org [19:12:13] Thank you nuria_ and HaeB [19:12:26] thanks leila. going to read the paper now. :-) [19:12:59] ragesoss: the other thing we tried to do is: let's forget about predicting motivations in general, let's try to build predictions as a function of the article the user is on. this path can be promising, and it can provide some use-cases, but again, for most of the product use-cases I could think of, asking the user would be way more accurate. This can be useful for editors, potentially, for example if we can give them a distribut [19:16:09] ragesoss: I'm a bit unfamiliar with the details of what entails for a taxonomy to be used as a set of persona. With this limited understanding: the way the taxonomy is built, it's pretty ground up and methodological, so I'd say yes: this is our taxonomy of Wikipedia readers and we should use it in other research about Wikipedia readers. Let me know if you read the paper and you disagree. :) [19:17:36] one thing I must say, ragesoss: we may do one expansion to the motivation question and that is adding something along the line of "I have edited this article before and am here to improve it." there are very few responses for this dimension, but it's something we're discussing within the team and with editors. It can make sense to add that given that editing is the fundamental activity on WP. :) [19:58:27] leila: now having read the paper... the taxonomy is very impressive. it maps well to intuition, despite being built ground-up starting from free text. I definitely hope we can start using it more widely in discussions and planning. :-) [19:59:11] ah! thanks for the feedback ragesoss. I'm printing it in bold/gold. :)