This development proposal is extremely welcome, and very exciting. I'm sure a number of Teahouse and Helpdesk volunteers will be willing to sign up as 'mentors' in any trial or roll-out. There are some really great ideas here.
I do apologise for the belated response. I have been collating comments to go into this reply elsewhere, whilst doing my best to assess how things run on Czech Wikipedia. I activated my 'Homepage' there, and by engaging with the mentor assigned to me. I used both a Windows PC and an Android tablet. Because of the language difficulty (and lack of browser translation) I have not yet tried on my iPhone in mobile view. I should say that I could not find a way to activate the Help Panel whilst dummy-editing a Czech article in either VE or Source Editor, so my comment on that particular Panel is based on just the graphic shown.)
Thank you for these thorough thoughts, Nick Moyes! And especially for trying the features out in the real world. You can activate the help panel on Czech Wikipedia in the "Editing" section of your preferences. As the Growth team spreads our features to more wikis, I hope that more people will try the features out before they're deployed. What do you think is the best way? Should we encourage users to visit wikis that already have the features? Or perhaps we could set something up on Test Wiki? -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
In terms of positive critical feedback, what few concerns I do have relate almost entirely to nomenclature. I do worry that the names given to certain new elements need to be carefully considered and possibly changed - not only by how new users perceive them, but also by how experienced editors will perceive them.
Your concern about nomenclature makes sense, and we do think about the right names to give things. One thing we'll need to be mindful of as we consider different names is what happens when those names get translated to other languages for other wikis. In other words, a term that may be perfect on English Wikipedia may not be as clear to other language users, and be more difficult to translate. I'll ask our engineers about how the nomenclature could be flexible across wikis. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
Medium concerns
Newcomer homepage. This is a great idea. Being able to optionally activate/deactivate the new Tab in 'Preferences' is very clever. I question the appropriateness of using the name "homepage" on en-wiki, and suggest that an alternative to 'Homepage' is needed. On en-wiki we currently two user-related tabs:
User page
Talk
...so, if there is now to be a third tab, named "Homepage"' I feel the name sounds so similar in purpose to WP:USERPAGE that it will lead to confusion by both new and experienced editors, alike. What alternative titles have been considered? Would "Start Page", "Newcomer's Page", "Task page", Welcome page or even "Ideas page" be more suitable? I might also suggest that in desktop view the Tab it could be positioned somewhat more centrally; i.e. keeping the pair of tabs for userpage/talkpage on the left upper side, with the read/edit/edit source tabs all to the right, and then placing the 'Homepage' Tab equidistantly between them.
Naming this page is definitely tricky. Beyond confusion with "User page", another possible confusion is with "Main Page". Interestingly, as we have spread the feature across wikis, different languages have translated this in ways that work for them. For instance, in Czech Wikipedia, the translation really means "Dashboard", because the translator judged that word to be more appropriate than the Czech word for "Homepage". One thing we want to be mindful of is giving it a name that will continue to be relevant throughout the user's journey. In other words, we don't want to use the word "Welcome" or "Newcomer" in the name because then the page may feel less useful once the user has been on the wiki for weeks or months. And to Sdkb's point above, the page could evolve as the user evolves, to become what an experienced user needs. One pattern we see in many software platforms is that each user has two personalized places: one place where you "broadcast" (i.e. your user page) and one place where you "consume" (i.e. your homepage). Some platforms have these all on one page together, with some parts only the user can see, and some that everyone can see. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
Can anyone else see the 'Homepage' Tab? It looks like it's a 'Special' page which only the individual user can see. Please could you confirm this in the documentation.
As currently designed, only you see your own homepage tab, and yes it is a Special page. We thought this felt right given that there are parts of the page that wouldn't really make sense for others to see, like the prompt to confirm your email or the option to ask your mentor a question. We felt that it was more like Preferences or Watchlist in that way. Do you think that's the way it should be? I clarified this point here in the documentation. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
AwardsThe Wikipedia Adventure has fifteen badges which a user receives for completing various tasks. Similarly, users receive a 'well done'notice after reaching certain numbers of edits (1,10,100 etc). Whilst not wanting to see an editor inundated with 'well done' messages, I might suggest a message within 24 hours of their first 'newcomer edits' made via this dashboard, noting how they'd contributed to specific tasks. A second message a week later could summarise their achievements made since then. A third message (say, a month after their first newcomer task edits' could be sent if no subsequent editing had been done. Its aim would be to send a "hey, we miss you!" message to show that they were valued.
Yes! This is the sort of "re-engagement campaign" that we were thinking of trying to do. We were actually considering using email for it (for users who have confirmed email addresses). The idea is that since we're trying to retain newcomers who, for the most part, have unfortunately left the wiki, we wouldn't be able to reach them on-wiki, and instead might have more success in their off-wiki email inboxes. What do you think about that? -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
Mentor this term is probably not currently well-understood on en-wiki - see essay WP:MENTOR - which has a bit of 'naughty' connotation here, whereas 'adopter' is a much more formal name used at WP:AAU, and is suggestive of longer-term support. I don't see why the current meaning behind 'mentor' on en-wiki can't be changed, though. But maybe 'Welcomer' might be a better alternative? It doesn't imply a long-term commitment from someone to follow the new editor around and guide them. My local Cathedral employs 'Welcomers' to answer simple questions and to orientate visitors. This term could be deployed on en-wiki so as to be less suggestive of someone who follows you right through your whole journey, but simply welcomes you in, answers a few initial questions and is there to help, if you need them. On Czech wiki I liked the clear fact that one editor (Janbery) had been assigned to me - it felt very personal. I was expecting the name to change when I returned, but it hadn't. Nice touch, that! (Czech mentor list) Maybe mentors could be provided with a list of new editors they've been assigned to, thus allowing them to view edits and maybe send a barnstar or two, should they be so inclined.
Regarding the name, I'll figure out with our engineers whether en-WP can adjust the word "mentor". We had a really interesting finding on this feature. In the first version of the homepage, there were two equal calls-to-action: one that let newcomers ask a question to the help desk, and one that let them ask a question to their mentor. Newcomers chose the mentor option almost every single time, and so we think there is something to this "personal touch" aspect. Your thought about mentors seeing who their mentees are -- I'm glad to hear you say this. Many mentors told us this at Wikimania, too, and we started to conceive of a "mentor dashboard" where they could see their mentees, who had been active, who had been reverted, and help keep a nurturing watch over them. What would you want to see on such a page? -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
I clicked on the edit help and found that I had been assigned a mentor. I instantly thought oh - what have I done as per @Nick Moyes comment " see essay WP:MENTOR - which has a bit of 'naughty' connotation here" So I raised a help desk and they were very nice https://en.wikipedia.org/wiki/Wikipedia:Help_desk#I_have_a_mentor_-_is_there_a_problem% The WP:Mentor essay has been updated to mention the growth project. It is pretty gloomy, and if I read the WP:Mentor page early enough I would run screaming into the hills Wakelamp d[@-@]b (talk) 15:44, 22 August 2021 (UTC)
Help panel: "Allows newcomers to ask a question directly to the help desk." The WP:HD is a bit technical and less friendly than WP:TH, so I'd suggest the Teahouse is the better first link.
I definitely agree -- each wiki can configure whichever page this feature should point to, and we would expect en-WP to slot the Teahouse in there. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
Default editor: Question to everyone: Should we point new users towards Visual editor or Source Editor? It does make sense to show new users how to edit with the simplest editing tool (WP:VE), but it must be made clear to them that it cannot and will not perform more complex tasks. New editors need to be aware of the benefits of learning how to use source editor if they choose to remain and do more than basic editing.
Workflow could we have a definition of what the WMF team mean by 'workflow' whenever it is first used in talk pages like this one? It's a team you guys use a lot, but we rarely do.
I'm sorry about that -- I try not to use too much product management jargon. When we product teams say "workflow", we're referring to a sequence of steps a user takes to accomplish a task. So the "suggested edits workflow" starts when a user first sees the option to do suggested edits and ends when they complete an edit. Does that make sense? -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
There's a tiny typo in this image file 'Gradually guides' should be changed to 'gradually guide
Hard or Harder. Which word sounds less daunting when describing the level of complexity of new editor tasks? Personally, I think 'Harder' sounds less extreme than 'Hard'.
Interesting question. One of the things we were thinking about in designing that moment is that we wanted users to understand that edits to Wikipedia can be all different shapes, sizes, and difficulties. And we wanted to communicate to the newcomer that the edits they may have in mind when they arrived (frequently, creating a new article) are probably too difficult for them to attempt right now. We want them to think, "Oh, creating a new article is hard. I better start with something easy." What do you think about this? -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
How is a new user alerted to their 'homepage'? Is it activated to begin with, or always, and can it be deactivated/removed in any way other than in 'Preferences'?
Newcomers find out about their homepage from a set of features we call the "discovery" features. Unfortunately, the only place that we have those documented is here in Phabricator. They consist of:
A clear button after account creation encouraging users to visit their homepage.
A popup pointing out where their homepage is (if they don't choose to follow the button).
A button on the "Contributions" page if the user has no contributions yet, encouraging them to visit their homepage to start contributing.
And when the user clicks to confirm their email address, the link now takes them to their homepage.
We recently did an analysis that showed that all those channels help newcomers find their homepage, and that depending on the wiki, between 40% and 70% of newcomers do visit their homepage soon after account creation. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
I often like to re-shape floating windows, so can the 'Help' panel be modified, or will it have just one fixed format?
We have heard this request a few times, and it is on our longer to-do list for the help panel. It's helpful to hear that it is also your preference. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
In the discussion above, Sdkb asked about removing templates. I agree that we can't expect new users to understand how and when it's OK to remove templates from the top of pages. That said, if they're directly addressing an inline template like [citation needed] it should be reasonable to advise them how to remove it. (I noticed on Czech wiki that few if any of the pages offered me for editing had big top-of-the-page notices).
Impact: - I don't have a problem knowing that 10,000 people have viewed an article on which I've simply added a comma or deleted a space; if I've genuinely helped improve the encyclopaedia with my teeny, tiny edit, then other people will still see my minuscule contribution and benefit from it. I don't think there's any other way of showing impact of a big edit, other then multiplying page views by the total change in bytes, but that would be a ridiculous thing to attempt. But, another suggestion: how about providing an additional link to a view recent traffic graph of, say, the last five articles a new user has edited? That could be additional visual encouragement? (Example on en-wiki; Example on cs-wiki))
The impact module does do that now! If you click the number of pageviews itself, it leads to the pageviews analysis tool for that article. I think my takeaway here is that it's not evident that that's what that link does, and so perhaps newcomers don't realize they can click it for deeper analysis. We'll think about that in future designs. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
Implementation I'd like to know more about how this exciting new feature will be implemented on large wikis like English Wikipedia, please. Because of concerns (expressed by Sdkb in their post, above) over matching new users to 'mentors' with a risk of swamping either individuals or the help desks, I assume there will first be some form of beta-testing, then a gradual rollout? I might be worried if it were suddenly implemented in one huge go - and I'm sure you folks are not planning to do that. I would hope that a small tranche of new editors would be given this functionality, and compared with a sample of new users who are not so the impact can be judged? I would certainly be happy to be added to a list of people either to trial the 'Homepage', or to volunteer as 'Mentors' - irrespective of what these elements end up being called here.
Great question -- we are also thinking a lot about the right way to scale this up, and that's why we're starting the conversation here early. One thing that will help is that we're deploying the features on increasingly large wikis over the course of the coming months. For instance, it is currently deployed on Arabic Wikipedia (which gets about 9,000 new accounts per month), and in May, we plan to deploy to French Wikipedia (which gets about 17,000 new accounts per month). This should help us learn about scale -- but English Wikipedia (which gets about 150,000 new accounts per month), is clearly a very different situation. I like the idea of deploying first to small groups of newcomers. We'll definitely be able to do that, which can help us identify any issues early on. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
Longer term, I could envisage that the 'Homepage' tab could evolve into a user-modifiable 'Newcomers' panel to offer something of greater use to the more experienced editor. There are a handful of really quite valuable 'Preferences' which are currently lost amongst the myriad of other settings at Special:Preferences (eg. 'Prompt me if I forget an edit summary'/'Enable Twinkle' etc) It could also become a place for an editor to compile a set of 'Quick links'of 'Favourite pages', perhaps offering the ability to define and store a set number of frequently used pages - be they Help pages, Policy Pages, Sandbox pages or articles currently being edited. But, as you rightly say, that's something for the future.
Live feedback Marshall: Have you considered in these 'lockdown times' of running something like a Zoom meeting to demonstrate an en-wiki version and gather verbal feedback from a handful of helpful, active editors interested and concerned with retention and welcoming? I'd be quite happy to participate if you felt there is value in doing so. Hearing one person say stuff, and then hearing a second person reply can often result in a third person coming up with some useful feedback thery m,ight not otherwise have done- and it's far easier than via this cold typing. Failing that, when you're ready, why not get yourself invited onto one of the hour-long recordings run by Fuzheado at Wikipedia Weekly Network. The Network covers a wide range of topics in quite some considerable depth, and this one would be a great opportunity to bring in Czech and other editors to explain how your initial rollout has gone, and what your future plans are for other Wikipedias.
I think this is a great idea for us to do at some point -- I agree that there is nothing quite like a live conversation for sorting things out and generating ideas. I didn't know about the Wikipedia Weekly Network. That looks really cool. I would prefer to have those larger conversations a big farther down the road -- perhaps a few months from now -- because I'd like to be able to speak to our experience on some of the other large wikis (like French Wikipedia), and because I would like our team to be up at full capacity (we're currently spread thin because of the pandemic). Would you be willing to help coordinate with us when it's time? -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
Agreed on all of these points/suggestions. Regarding the question of whether to use VisualEditor or the Source Editor, that's a hugely loaded question given VisualEditor's very charged history at en-WP, so that's something to step very carefully around, and to ideally break off into a distinct separate conversation, so that opposition to using one editor or the other doesn't bleed over into more general opposition. {{u|Sdkb}}talk06:42, 21 April 2020 (UTC)
Sdkb and Nick Moyes -- thanks for bringing up this question of which of the two editors newcomers are "nudged" toward by the features. We thought that different communities might have different preferences here, and so we've built the feature such that each wiki can configure which editor opens by default when newcomers begin doing suggested edits (the user would be able to switch). So far, all the communities we've worked with have preferred Visual Editor (Czech, Arabic, Korean, Vietnamese Wikipedias). When it is time for us prepare these features for English Wikipedia, I think we'll be able to implement the community's preference here for these open-ended sorts of edits, and maybe we'll have some data from the preceding wikis on how their choices are working out. I think Nick Moyes makes a good point about the simultaneous need to (a) give the user a good first experience so that they'll succeed and want to return, while (b) also making sure they know about the source editor and its capabilities. I also want to mention that we're in the early phases of planning a project called "structured tasks". The idea is to hold newcomers' hands as they do their first edits, so that it is clear to them how to get through the several steps of simple edits (e.g. click "Edit", highlight some text, click the link icon, choose the right article, click publish...) I'll return with more information about "structured tasks", because I think that this question about Visual Editor vs. source editor will be relevant there, as well. -- MMiller (WMF) (talk) 01:07, 29 April 2020 (UTC)
How to write a good article
The "General editing help" tab has a link called "How to write a good article", but it links to the entirety of the Manual of Style, an absolutely behemoth document that even I don't know cover to cover. For starters, it should not be the first link in the list, it should probably come last. Secondly, it should link to a more concise page, an essay if it has to. CaptainEekEdits Ho Cap'n!⚓18:51, 7 May 2021 (UTC)
Several of the items do link to pages in the Help:Intro series, though.
Regarding the MoS, Help:Introduction to the Manual of Style/All would be a simpler alternative to the full MoS, which I agree is unwieldy. It's impossible to be both comprehensive and instructive to newcomers in the same place, which is why simplified pages are so important (the main problem being that there are tons of competing "simple" versions that aren't actually all that simple).
More generally, I've been paying attention to the language used in the help tab. Most of it is good, and I sympathize that whoever wrote it had a tricky task trying to be quick and concise but to not oversimplify. But a few things irk me a bit, like the fact that tab 5 of the references instructions says "Reliable sources include books, news organizations, and magazine articles." Not all media are reliable, so it'd go a long way to change it to "Reliable sources include reputable books,..." Also, there's no way the "how to create an article" link should go to the retired Wikipedia:Article wizard/version1, rather than the modern article wizard or Help:Your first article.
I hope that the help links text is eventually stored somewhere on-wiki, where the community will be able to modify it as our help resources evolve. {{u|Sdkb}}talk19:38, 7 May 2021 (UTC)
If the aim seems to be to stop new users leaving in frustration, the first thing should be don't create a page unless you talk to your mentor Creating a page and getting rapid RfD, then rage quitting is common based on my friends experience.Wakelamp d[@-@]b (talk) 15:50, 22 August 2021 (UTC)
More?
I seem to be getting more mentor questions in August. Is the feature being shown to a larger portion of new users, or is this just random luck? ⁓ Pelagic ( messages ) 19:59, 9 August 2021 (UTC)
Hi @Pelagic -- we haven't changed anything about the setup or the number of users getting Growth features. I also checked and saw that the mentor list hasn't changed (e.g. if half the mentors suddenly quit, the remaining mentors could expect twice the questions). So yes, I think an increase in questions is just luck of the draw. How many have you been getting? How has the quality been? MMiller (WMF) (talk) 05:05, 13 August 2021 (UTC)
My question count is slightly down, so I think it's the luck of the draw @Pelagic: - obviously we get the same amount of mentees, but whose mentees ask questions will fluctuate. On the plus side, the quality is slightly up. Nosebagbear (talk) 08:43, 13 August 2021 (UTC)
Comments from the African Wikimedian Community
Hello colleagues from the African community who are supporting the deployment of Growth Features on English Wikipedia please look at the sections below and comment where necessary. You can either show support on behalf of your affiliate after consulting them, or yourself. Don't forget to add your signature. sandioosesTextMe19:19, 4 May 2021 (UTC)
I fully support the deployment of the new Growth Features on English Wikipedia. It's going to go a long way in helping to guide new editors in a very timely process. Owula kpakpo (talk) 20:25, 4 May 2021 (UTC)
Yes! As a community organizer and new editor recruiter, I support the development of these features on the English Wikipedia. The Living lovetalk10:23, 5 May 2021 (UTC)
Hello Sandiooses, Owula_kpakpo, and Em-mustapha! I wanted to get back in touch with you as the trial of the Growth features has reached a new stage. The initial trial for 2% of new accounts has gone well enough that community members have agreed that we should propose to now try the Growth features with 25% of new accounts. I'd like to invite you to check out the proposal on Village Pump and weigh in if you want! Thank you! -- MMiller (WMF) (talk) 23:54, 2 September 2021 (UTC)
Mentorship thoughts
So I just finished my first interaction with a mentee, and while it's only a sample size of one, I wanted to note down some initial thoughts.
Overall, the interaction went very well—the editor was seeking help with an article that had been draftified, and I was able to guide them to getting it back to mainspace and they seemed very appreciative. It's impossible to tell whether that's selection bias or a fluke or not, but it's a small positive sign, as many similar interactions at the Teahouse don't get to that outcome.
The main thing that stood out about the interaction to me was that it was solely two-party, compared to the environment at the Teahouse where we all collaborate to answer questions as a group. This certainly made me feel more invested in their situation and inclined to follow up on it, as I knew if I didn't help them out there was no one else who would step in. That has positive elements, but there are also downsides: if I hadn't been available, the response could've taken longer, and if I'd given bad advice, there would've been no one to catch me on it. I do think a situation sometimes happens at the Teahouse where someone asks a simple question and gets overwhelmed with a bunch of responses saying similar but slightly different things, so it may be helpful to have this model resulting in singular, more concise answers. Back to the downsides, the flipside of feeling investment in/a responsibility to the situation is that I think mentors may feel pressure to answer even when they don't feel like it, which could contribute over time to resentment for unpaid labor if mentors aren't adequately recognized (small tokens like Meta:Merchandise giveaways go a long way).
An aspect of the interaction that felt a little weird was that it took place on my talk page rather than the other editor's. I think it'd be better to have it take place at their page, as their own page is easier for them to find, and their question has more to do with them than it does with me. If the volume of questions was a lot higher, I could see potential mentors being discouraged from signing up because they don't want their talk page to become clogged up. {{u|Sdkb}}talk23:34, 12 May 2021 (UTC)
Great observations. Are there any plans to build a rate limit into things, i.e. at WP:NPPS we can indicated how many people we can help at once? I'm guessing the WMF team will have more to say but I like that the talk page messages are on the mentor's talk. One because the mentor is more likely to have talk page watchers so some sort of communal response may still be possible and two because this way there's no missing a notification for help. Best, Barkeep49 (talk) 00:14, 13 May 2021 (UTC)
I don't see why this would be a problem; most of the question threads so far only have about 3 or 4 messages. For the test deployment, mentors should be fine, as we already have ~10 mentors. Not sure about after test deployment. The best I can think of is making a wikipedia ad. Sungodtemple (talk) 12:26, 13 May 2021 (UTC)
Thank you for sharing these thoughts with us, @Sdkb.
To be honest, there is no perfect system regarding mentoring. We think there is one that works better, though. You highlighted most pros and cons regarding using a central place (like the Teahouse) or the mentor's talk page. An interaction with an individual is less complicated to apprehend for most new users, hence we privileged it. Don't forget that newbies are impressed by Wikipedia, so it is easier to contact someone that reaches at you directly and personally, than coming to a share page where you're not sure of being legitimate to post there.
The feeling of unpaid labor you mention is not something I heard about at other communities. I think mentoring is considered as a new way to get involved on Wikipedia. One would fix wikitext, an other would patrol recent changes, a third one will categorize. Mentoring is a new way to participate, it just needs to be recognized as such. For instance, at my home Wikipedia, mentoring is a great plus in your favor when you run for adminship.
Regarding the specific point of replying on your talk page, Barkeep49 gave two of the reasons why we encourage mentors to reply on their talk page (and ping newbies). The third one is that it will be very difficult to scale when you will have interecated with +100 users. Even if their activity may not be constant, it means a lot of noise in your watchlist. And I speak from experience there. :)
Marshall will soon share with you the schedule regarding the test. You could consider to have this call for volunteers when the test will be over, Sungodtemple. :)
Thanks for writing down your thoughts on the mentorship interaction, @Sdkb. I'm glad your first mentor situation was so productive -- your proactive approach to working with the mentee is what we think can work really well. I will say, though, that I expect many mentor interactions to be less satisfying. For instance, it's common for a mentee to ask a question, the mentor to answer, and then to not hear back from the mentee (frequently because that mentee has already left the wiki).
You're picking up an important balance that we've considered: speed vs. personalization. If newcomers ask questions to a central place like a help desk or Teahouse, then they'll get an answer quicker, because there are many watchers. But if the question goes to a mentor, there is an opportunity to make a human connection that could be more comfortable and inspiring for the newcomer. When we first deployed the homepage in 2019, it had two options for asking questions: "ask your mentor" and "ask the help desk". The buttons were the exact same shape and size, but newcomers chose the mentor option about 10 times more frequently. This is what made the decision for us -- it seemed like newcomers felt more comfortable with mentors than with the help desk style situation.
It's an interesting idea about having the questions go on the mentee's talk page instead of the mentor's talk page. We put them on the mentor's talk page so that it's as simple as possible for the mentor to notice and answer them -- they don't have to go check some other place; it's already an inbox they expect to go through. But perhaps the questions could go to the mentee's page and automatically ping the mentor? I think a downside to that might be the weirdness (that Wikipedians are used to) of asking someone else a question on your page instead of their page. That might confuse newcomers.
I definitely like the idea of rewarding mentors. While mentors can always remove themselves from the list if they feel burned out, showing them their value is important. One small thing that happens on Arabic Wikipedia is that mentors can display a userbox. You can see it at the bottom of this user's page, with the Growth team's "W" logo.
@Barkeep49 -- about your idea of rate limiting, we've learned a bit about how the dynamics of mentoring tend to work. Newcomers are always flowing into the wiki, and they each get randomly assigned to a mentor. Then a small percentage of the newcomers ask a question, and small percentage of those continue to correspond with their mentor and ask questions. If we calibrate the number of mentors correctly, then even though newcomers are always flowing in, it lines up with the tail end of the previous newcomers tapering off their questions, and stays like a constant flow. Does that make sense? MMiller (WMF) (talk) 21:11, 14 May 2021 (UTC)
I was interested that the help desk did not get as many clicks. Mentor has a sort of concierge feel about it. I am not sure whether Wikipedia does A/B testing, but it would be interesting if it you added the words Priority help desk (2 hours) and a Happy Wikipedia logo, and Mentor (up to 3 days) Wakelamp d[@-@]b (talk) 15:55, 22 August 2021 (UTC)
Following up here, I have to say that the point about mentors feeling unwanted pressure to respond has continued to hold up. There's an aspect of being a mentor that feels distinctly more burdensome than being a Teahouse host, and it emanates mainly from the fact that at the Teahouse, I can choose which posts I respond to, whereas as a mentor, I don't get that choice. I've encountered many posts from paid/COI/CIR editors that aren't so awful as to be ignorable or warrant reversion, but where I would also skip over them if I encountered them at the Teahouse, since I prefer to help out newcomers with the potential to become experienced editors. The lack of choice makes the difference between feeling like a volunteer and feeling like an unpaid customer service representative. Perhaps others are feeling differently and don't mind, but as this feature scales up, if the number of mentor signups isn't as expected, you may want to explore ways to address this problem. Some things you could consider are an option for mentors to indicate "I don't want to respond to this question, please reassign this editor to a different mentor/the help desk", and limiting the use of the feature by COI/paid editors (who have an obligation to thoroughly research issues they encounter before asking for help, rather than abusing the volunteer nature of the project by asking editors to essentially do their work for them). {{u|Sdkb}}talk05:13, 6 September 2021 (UTC)
The lack of choice makes the difference between feeling like a volunteer and feeling like an unpaid customer service representative. — I've always felt more like an unpaid customer service representative when communicating with newbies, whether political POV pushers on articles or COI editors at AFC. I don't think the mentor panel is doing anything other than shifting existing antagonism to a different location (my talk page itself). COI editors already abuse our site and volunteers by their actions, just normally at different venues. You might consider keeping a form response in a sandbox somewhere for COI questions. Personally, I have found the mentor tool to be very helpful in COI cases: I've perhaps discouraged a few NOTHERE people with my answers (not intentionally antagonistic, just frankly honest that we don't like COI editing) that may otherwise have wasted a lot of AFC time. — Bilorv (talk) 14:42, 11 September 2021 (UTC)
I sympathize with the desire to focus on new editors who demonstrate potential to improve, as I have the same inclination. But I think it's better overall for the mentoring initiative if the workload for less-promising editors is shared more broadly. Providing a venue where mentors can swap their assigned newcomers might be useful—it can allow for better matching of interests and further balancing of workload, should a mentor become overloaded. isaacl (talk) 15:08, 11 September 2021 (UTC)
Sdkb, This is an interesting thought. Especially while we're not rolling out the mentorship option to all editors, it might make sense to give random potentially promising newbies rather than continuing to waste mentorship slots on people who are shown not to be promising editors. In theory, I'd support having a button that de-enrolls someone from the mentorship option and gives the opportunity to someone else. There could conceivably be abusive de-enrollments by bad mentors, but I imagine that there would be very few of these relative to the good de-enrollments that free up slots for others. Calliopejen1 (talk) 18:57, 12 September 2021 (UTC)
VPR Sign-off/Next Stage
Hi @MMiller (WMF):, just checking that you'd seen that you'd gained a unanimous sign-off over at VPR for the next stage of roll-out. Only additional aspect of the close was basically to cross-post the next set of outcomes data from here to the village pump once you've assessed it.
We do still need some additional mentors added to the list, as we've only added a few thus far (though we probably can deal with a higher q density than we've had, at least another 5 would be good). Nosebagbear (talk) 13:35, 13 September 2021 (UTC)
Hi @Nosebagbear -- yes, I saw the sign-off! Thank you for helping us get to that point. It was good to see all the support. I just created this Phabricator task to make the change a week from today, on September 20. That should give us enough time to get some more mentors signed up. I'll post a message for the Teahouse and for Women in Red.
We made the change this morning, and we're now giving the Growth features to 25% of newcomers, with only 5% getting mentorship. Thank you everyone for helping push along to this exciting step! There are now 33 mentors signed up, which I believe will be enough for this stage. I'll keep an eye on Recent Changes to see what kind of volume comes in and if any problems crop up. But other than that, our next step would be to generate the metrics and continue the discussion in about a month. Nosebagbear -- do you recommend that I announce this elsewhere? The VPR discussion is closed, so I don't think I'm supposed to change that. -- MMiller (WMF) (talk) 16:55, 20 September 2021 (UTC)
MMiller (WMF), I would've suggested notifying at WT:TH but it appears that a lot of the hosts are already here or the rest of the project is not actively following. Maybe bring it up to the folks at the Help Desk? I think a formal proposal will garner some more attention there. Panini!🥪17:11, 20 September 2021 (UTC)
Removing inactive members more rapidly than we might otherwise
Hi all,
As we head towards what will be a large mentor list, Xaosflux's bot note above made me realise that this Mentor list should have a much more aggressive pruning of inactive members than we do for other userrights.
This is because while an inactive admin might be bad for the project as a whole, it doesn't cause direct damage to specific individuals.
However, if a mentor was on the list, but inactive, for 3 months, they might get (say) 30 new individuals asking questions with no chance of getting a suitable response and as the big flaw we've discussed with the mentor system is the "any error by a mentor won't be seen by someone else", this lack of answer would mean a complete lack of answer.
As such, I'd like to see if there would be support for pruning the list far more actively than we would usually - 1 month being my proposed time. Mentors who returned after that time would of course be welcome and encouraged to re-add themselves, and I can see dropping off (hopefully self-removed, but often not) and back in being quite common, but it would reduce the problems. Nosebagbear (talk) 11:09, 21 September 2021 (UTC)
Nosebagbear, I think that's a fair concept. I do also propose of posting a template on their talk page, explaining why they've been removed and encourage them to re-add themselves to the list when they become active again. Panini!🥪11:22, 21 September 2021 (UTC)
I wonder if there's a way this can be more dynamic than proposed - i.e. the software itself doesn't send messages to users who haven't edited in the last [suitable time frame], but there's no need to actively remove editors from the list or have them re-request being added? Sam Walton (talk) 11:30, 21 September 2021 (UTC)
If that's possible, that would be even better, of course - I suspect returning individuals would continue to be happy to answer questions but may not always remember to rejoin as mentors. However, unless it's very simple on the software side (at least it could work off global contributions so it will be project agnostic), I'm not sure we'd want to divert the team away from some of the other bits they're working on that we want. Nosebagbear (talk) 11:34, 21 September 2021 (UTC)
A related thought: It'd probably be possible to make an automatic list of threads from newcomers that have not received a reply. That might help ensure everyone gets a reply and/or identify inactive mentors (including those who stop mentoring but are still editing elsewhere). However, the challenge would be to prevent it from filling up with all the very low-quality posts that justifiably don't get a reply. {{u|Sdkb}}talk13:21, 21 September 2021 (UTC)
On top of that, questions get answered in different locations - to avoid issues, I sometimes override my normal "respond where first asked" and go over to their Talk Page. This is especially true for the Hi-style "questions". Nosebagbear (talk) 14:00, 21 September 2021 (UTC)
I'm glad everyone is talking about this! This is definitely a challenge with mentorship that we've seen in other Wikipedias: that mentors either stop being active in the wiki, or that they stop answering their mentees (perhaps because they are bored with or burned out with mentorship). Sometimes, community members have clicked through the mentor talk pages manually to find and remove those mentors who stopped participating (in most wikis there are fewer than 15 mentors, so it doesn't take long). But obviously, that's not a scalable solution. I think the ideas that have been mentioned above could all potentially be good -- ultimately, I think it should be up to each community how to manage their mentors (and we want to supply you with the tools to do it).
Sdkb mentioned the idea of noting whether a mentor has many unanswered questions, and yes, the hard part is parsing the talk pages to see if there's a reply. It's something that Martin Urbanec (WMF) has thought about, and maybe he could fill us in on how possible it might be. -- MMiller (WMF) (talk) 03:36, 22 September 2021 (UTC)
The easy part is to get sections titles for the questions -- then, it should be only a matter of signature-detecting (if the section has only one signature, it is likely an unanswered question, if it has multiple, it is likely answered). I didn't yet start any actual work on this (eventually, I will, as part of engineering the mentor dashboard), so those are just high-level ideas about approaches that can be used. Other possible approach (although only applicable from within MediaWiki) is to use DiscussionTools functionality to detect who talked in the given section, and go from there. Best, --Martin Urbanec (WMF) (talk) 07:55, 22 September 2021 (UTC)
Just as MMiller (WMF), I'm also happy everyone's talking about this. I would however like to put a few notes here, that can be useful when working on a bot to curate this list. First of all, if you intend to reassign newcomers to someone else after removing an user from the list, it is a slightly disruptive acton (for the small subgroup of the mentees which are currently active on the wiki), as it both logs to the logs, and causes an echo notification for the mentees. You could theoretically just remove the user from the list – experience shows mentees tend to ask questions in the first 24 hours after registration, and since the removal results in no more mentees assigned to them, it should be enough to significantly decrease the number of questions the user receives. To be fully sure, you'd need to reassign all mentees assigned to a given mentor.
The reassignment is not only visible to newcomers, as described above, but it is also not really immediate. We currently assign a mentor to everyone, including no-Growth users and users with mentorship disabled -- that is because some wikis wanted to merge their own mentorship system with Growth's, so we introduced the #mentor function, which allows you to see someone's mentor. To satisfy curiosity, here's an anonymized list of all current enwiki mentors by number of mentees: phab:P17313 (note I mentor myself for test reasons, and Xaoflux was testing the API for setting mentors with their alt accounts, so they also have a single-mentee mentor, cf [1]).
I would like to ask the community to take the factors described above into consideration. As always, let me know if you have any mentorship-related questions, I will be happy to help. Best, --Martin Urbanec (WMF) (talk) 07:55, 22 September 2021 (UTC)
We started giving the Growth features to 2% of new accounts on June 8, and we now have a month of data to look at from this test. I think we should use this information to do these two things:
Decide whether to proceed to an increased percentage of newcomers (perhaps 25%, as suggested by Nosebagbear).
Brainstorm any ideas for how the wiki can better handle increased activity coming in through these features.
Below are the results from looking at data from the 30 days of June 8 to July 8, along with what we might expect them to be each month if eventually deployed to 100% of newcomers. That number is just calculated by multiplying by 50. I also included some of my own reflections on what we're seeing, but my reflections are not as important as those from the rest of the community, so please weigh in and disagree!
Suggested edits
Edits
381 suggested edits were completed. At 100%, we might expect this to be 19,050 per month. In the month of June 2021, newcomers made about 160,000 edits, so a potential increase of 19,000 does not seem like it would be a disruptively large change (plus not all these edits would be marginal increases; some replace edits newcomers would have otherwise made on their own).
34 of those 381 edits were reverted. That's a revert rate of 9%. During June 2021, the overall revert rate of newcomer edits was 27%, so these suggested edits seem to be of consistently higher quality.
Users
68 distinct users completed suggested edits. At 100%, we might expect this to be 3,400 per month.
10 users completed suggested edits on three or more separate days (15%). This shows that a good portion of the users were engaged enough with the feature to return and continue with it after their first session.
36 users completed three or more suggested edits (53%).
10 users completed ten or more suggested edits (15%).
Observations
In my own observations of the diffs coming through with "#Newcomer task", I have seen lots of good-faith edits that add links, clean up grammar, and sometimes add references. This link filters Recent Changes to just those edits.
A couple days after the 30-day window of analysis ended, a user completed 460 suggested edits over about three days. These edits, however, occurred on only about 25 articles, with the user saving dozens of separate edits on some of them. These seem to have amounted to constructive changes, but they were saving after every word or link change, as opposed to saving all their changes at once.
We received some qualitative feedback from a newcomer here, describing positive experiences with the features.
Mentorship
Edits
106 mentor questions were asked. At 100%, we might expect this to be 5,300 per month.
7 of those 106 questions were reverted. That's a revert rate of 7%.
Users
92 distinct users asked questions. At 100%, we might expect this to be 4,600 per month.
0 users asked questions on three or more separate days.
2 users asked three or more questions (2%).
Observations
This link filters Recent Changes to just these mentor questions.
Users tended to ask only one mentorship question, and not to continue to ask more questions via the mentorship feature. This may be because they only had one question, because they continued conversations on the mentor's talk page in the conventional way, or because they never saw the mentor's response (perhaps they did not register with an email address and did not return to the wiki to check for a response).
For qualitative thoughts on the types of questions being asked, I'll ask the mentors to reply to this thread.
Discussion
In terms of suggested edits, it seems like things are going well -- newcomers are making edits that are reverted at low rates. Most of them do more than one suggested edit, and many come back on multiple days to do more. But they are not making so many edits that expanding to 100% of newcomers seems like it would overwhelm patrollers.
Mentorship is where there are bigger potential challenges, because of the number of mentors that would be required (a discussion that we began in this section above). If we were to be receiving 5,300 mentor questions per month, that would be about 175 per day. If we think that mentors might want to only receive one question per day, we would need something like 175 mentors. I think this question is where we need the most ideas and opinions.
How many questions might mentors be comfortable receiving per week?
How big of a group of mentors would be manageable? Would the English community want to have any sort of process for signing up, or for peer-monitoring the work of mentors?
What ideas can we think of for assembling and managing a large group of mentors? Here are some that I've been thinking about:
Perhaps there are volunteers who want to get a high volume of questions, and would be happy to spend a lot of their wiki time on answering them. I could imagine making it possible for these people to sign up to get a "high" number of mentees, while other mentors might prefer to sign up for "medium" or "low".
Perhaps rather than having one big pool with hundreds of mentors, mentors could be grouped by topic or geography (e.g. Sports, History, Science, South Asia, etc.) And if we encourage newcomers to indicate their topics of interest, they could get assigned mentors from that topic. Those topical groups of mentors might only have a dozen or so people in them, making them easier to manage amongst themselves.
Thank you all for reading and following along. Please speak up with what you think about what we're seeing here, what you think the right next steps should be, and ask any questions you have! MMiller (WMF) (talk) 18:48, 14 July 2021 (UTC)
The data, qualitatively and quantitatively, all looks promising, except, as noted, the ability of mentorship to scale. Even the 460 edit spree feels like something either done in good faith, in ignorance, or in a desire to inflate their edit count none of which would be beyond the realm of enwiki to help. As for the mentoring issue, I'm wondering if this would be something that could be tied into the Volunteer Response Team (nee OTRS). Those editors are already screened which strikes me as very helpful so we're not leading new editors astray. The screen is actually beyond what we would need for mentors but has the virtue of being an existing process. But the OTRS backlog also suggests it can't just be a 1:1 thing. I think NBB's idea of a more public posting - at some village pump - ahead of a 25% trial feels like a good next step to getting more mentors, more thoughts, and more data about how this might scale. Best, Barkeep49 (talk) 18:57, 14 July 2021 (UTC)
Yeah, the 460 edits is unusual but if it's an editcountitis thing then that's not really to do with the mentorship. I wouldn't be surprised if it's a tendentious sock looking for extended confirmed rights to mess up a political article, but that wouldn't be mentorship-related and it's not the first time I've seen it. Weirder and more inexplicable things with newbies happen every day. — Bilorv (talk) 19:45, 14 July 2021 (UTC)
My very first thoughts are:Editors should not be shown the mentorship resource before they've made a few constructive edits. A sock's first edits were sending me questions via the module. If that happens on a regular basis, I might considering opting out. And there seem to be editors who sign up, ask a question, and never come back for the replies or to edit. I am not convinced that answering questions from throwaway test accounts is good use of mentor time. On an unrelated note, I had an editor come to my talkpage to tell me not to contact them. The message was not tagged but I can't think of any other reason they would think I would contact them. So, it perhaps needs to be clear/er that it is a private recommendation, and only a mentee can disclose the relationship and initiate contact?Best, Usedtobecool☎️19:09, 14 July 2021 (UTC)
I would rather that editors be shown mentorship resources immediately. The way I see it is that as a mentor, you need to be willing for all of the time you put in to answering questions to potentially be "wasted" (not that you can ever actually tell the effect that it had). Anyone familiar with answering Teahouse questions, declining AFC drafts etc. should know this feeling: your time and effort goes into a vacuum and you will often get no replies, no sign of activity, no evidence that anyone has noticed that you exist. Time wasted dealing with socks or time wasted with hopeless newbies is not necessarily increasing, but just being redirected towards this mentorship scheme, because socks use every available attack vector and a CIR case does too, and the pool of people who give up immediately and never return also expend our time in various ways. Instead of a mentorship question, they could be making a test edit or a Teahouse question or whatever, but there's still wasted time.
If you sign up to mentorship, you should be prepared for this, but you can find a workflow to moderate its effects. I've been meaning to write an essay sometime about proof of work applied to Wikipedia activities. Here, maybe allocate 3 minutes to answer a question (or 90 seconds if you want) and make sure you don't spend longer answering it. End it with "is there anything you want me to go into more detail about?" If you get a response (proof of work by the newbie) then you go into more detail. If not, don't. This way you don't go all out with a detailed answer that's never seen. — Bilorv (talk) 19:45, 14 July 2021 (UTC)
Hi @Bilorv -- I like that "proof of work" concept! Perhaps we could incorporate that into some written guidelines for mentors, once we have this feature fully up and running. MMiller (WMF) (talk) 03:36, 23 July 2021 (UTC)
Hi @Usedtobecool -- in thinking about the experience you had with the mentee telling you not to contact them, my guess is that they saw your default "greeting" in their module and assumed that you had proactively sent it to them (like in the screenshot at the right). We thought that these personal "greeting" made mentors seem more accessible, but perhaps they confuse some users. What do you think? Mentorship module with greetingMMiller (WMF) (talk) 03:35, 23 July 2021 (UTC)
That must be it, @MMiller (WMF)! It does look as though the mentor has sent them a message, and mentors wouldn't necessarily have known to design their note to avoid confusion. So, heads up to the mentors and probably a header or footer to the quote to the effect of "Mentorship philosophy" (but smarter) might work (don't know if it's worth the space with just our sample of 1, but something worth discussing in the next redesign meeting). Best, Usedtobecool☎️04:53, 23 July 2021 (UTC)
The data is quite promising. If we could increase the number of new edits by 10-15%, and with a 90% quality rate, that would be a fantastic success. Assuming that more of those editors would be likely to stick around, it could increase our overall number and quality of edits over the long run. If we had even 10% more editors sticking around, that's probably 10% more admin candidates, 10% more featured articles, and so on. Wonderful work!!
With regards to mentor questions, I don't think we would struggle with having questions answered. We already have a very full slate of helpers at the teahouse and helpdesk, such that a question rarely goes more than 10 minutes before being answered. I think that speaks to the fact that we could process many more questions per day as a community. The process merely needs to be publicized. I agree that the next step here should be to put this forward to the community at a broad forum, posited as a request for input ahead of rolling it out to 25% of newcomers. This should be followed up by another community consult before then loosing it on 100% of newbies. You folks are really doing a great job, I think this is perhaps the single most important WMF initiative. CaptainEekEdits Ho Cap'n!⚓19:36, 14 July 2021 (UTC)
While it's hard to say what the most important of any WMF project is, I do feel that though most are aware that our editor number issues are all down to the abysmal retention, it doesn't seem to have been matched by a realisation that if Growth can provide any significant retention bump at all that will have literally monumental effects. A huge example of how a project doesn't have to be flashy to provide major benefits. Nosebagbear (talk) 19:59, 14 July 2021 (UTC)
The thing about teahouse and helpdesk is that they are answered by two or three people. If one is wrong, then it's pretty clear that they're wrong. In mentorship only one person answers the question, which means they have to be even more reliable. Sungodtemple (talk) 13:43, 15 July 2021 (UTC)
I think scaling up to 25% is the right next step and the most major priority is working out how to assign different proportions of mentees to different mentors. — Bilorv (talk) 19:45, 14 July 2021 (UTC)
(edit conflict)Evaluating my own interactions I have had questions from 4 users. One "Hello Everyone" from a users whose 2 other edit don't indicate they were here to contribute. 2 questions were from users who only have edits to my talk pages. The last question one was "how approve page" from a user trying to get a page for their tiny YouTube channel. I think Usedtobecool's filtering suggestion would be beneficial in both decreasing work load and number of non-productive questions. My cut off would be something like 10 edits before assigning a mentor (no time requirement so not quite EC). I think we can handle a 25% trial and would support starting that quite soon. Are 2% of newcomers still getting the tools or not btw? If and when we deploy the tools in full I feel we should have a one year evaluation on it. With that much data we could see if there are parts that aren't working and adjust accordingly. --Trialpears (talk) 19:49, 14 July 2021 (UTC)
Hi @Trialpears -- thanks for relating the experience you had as a mentor. Yes, 2% of newcomers are continuing to get the feature. I was thinking we could leave the setting at 2% until we upgrade it to a higher percent, allowing us to continue to have a trickle of fresh data as we consider next steps. Do you think that's the right approach? MMiller (WMF) (talk) 03:43, 23 July 2021 (UTC)
The questions I had posted to my page were generally extremely low quality (much worse than what the TH receives). Three non-question hellos from two users. Two bizarre non-WP questions from a single user who may be mentally ill (?). One "how do i work here" from someone who didn't make any other edits (not clear whether they wanted to contribute or what the question meant). Four questions from three users attempting to use WP for advertising. One pretty good question from someone who wanted to write a biography of their professor who is supposedly an animation pioneer. They never followed through but I provided information that could help them. Not sure if this is representative. I'm wondering if there is some way to improve the interface to discourage people from non-WP related posts/questions, which were about half of what I got. Calliopejen1 (talk) 20:13, 14 July 2021 (UTC)
Thanks for summarizing your experience, @Calliopejen1. Maybe there are things we can do to the copy in the mentorship module to help people keep their questions productive. Right now, the language in there says, "Ask your mentor a question", and then the prompt is, "Say hello and ask your question". It's possible that the "say hello" part causes those "non-question hellos" that you received. We do want the experience to be inviting, but perhaps there's a way to accomplish both objectives better. Do you have ideas for how it could be different? One thing I can think of is that we could show examples of good questions, so that users get a sense of the sort of things they can ask. MMiller (WMF) (talk) 03:46, 23 July 2021 (UTC)
@MMiller (WMF): Maybe keep the first the same (or "Ask your mentor a question about editing Wikipedia") and and change the second to "Ask your question about editing Wikipedia"? Not sure if this would really make that much of a difference, but might be worth a try. Calliopejen1 (talk) 15:28, 23 July 2021 (UTC)
I don't know whether there are 175 users who can answer a question every day. Perhaps on average in the sense that there's probably more people that can answer a few questions every few days, but a lot of users don't log in daily or may not want to deal with questions daily. I like the idea about letting people control the volume they'll get. Topical grouping is also a good idea, and may lead to more interesting questions for the mentors. As for signing up, a page protected to extended-confirmed editors (min 500 edits and 30 days registered) seems fine long-term IMO. I don't know of any systematic peer-monitoring processes for editors; usually problems are reported if someone happens to stumble onto them. ProcrastinatingReader (talk) 20:37, 14 July 2021 (UTC)
Adding: I agree with notifying a village pump before scaling. In regards to volume: it's probably a good idea to implement preferences relating to volume first. If no mentors with 'remaining volume' are available (eg too many questions have been asked on the day) then I think it would be best to cut off new mentor questions and replace the button with a link to the WP:Teahouse or something. That's to prevent mentors being overwhelmed and removing themselves from the list. ProcrastinatingReader (talk) 20:42, 14 July 2021 (UTC)
Thanks for the update and info! I concur that we're good to jump to 25% or similar. Regarding thoughts on mentoring, I don't have much to add to my initial thoughts from above; I haven't been getting too many questions. {{u|Sdkb}}talk21:16, 14 July 2021 (UTC)
To answer your questions: 1) Thinking for myself, I would personally like about 1 a day. However, since most of my questions have been "hi" or "what do i do" 2 or 3 shouldn't be a problem. 2) I think there should be some sort of sign-up system for becoming a mentor. The Teahouse has sign-up requirements but some inexperienced editors could bypass it easily (with good intentions of course). Maybe something like the AFC sign-up? 3) I love both ideas! I was gonna suggest people being able to choose how many mentees they want per week if you didn't. I was hoping something like the second solution would be put in place. Thinking about myself again, I have experience in the video game field and that's it. If someone were to ask me about dementia with Lewy bodies I wouldn't be able to help them as easily. Ooh! What if they were allowed to pick from a group of users who specialize in something from a broad category, and it lists out what each user specializes in within that category? From there, they could search the "Technology section" and find users who specialize in HAM radio tech, personal computers, and pick me out for a video game related question. We would have small blurbs that say what we know about (eg. "Panini!: Video games"). I think that's too complicated of an implementation at this point. I'm rambling; I like what I see! Panini!🥪22:50, 14 July 2021 (UTC)
Personally, I had a not-so-good experience with newcomers. Just look at my talk page. I could definitely handle 5x the questions. Also I feel like mentors have too much power. They can choose who their mentees are. This means that an EC bad-faith editor could just look at the list of newcomers and automatically assign the newcomers to themselves. This means the one bad-faith editor could very easily mess up the mentorship module. Maybe there should be a user right, like Panini! mentioned above. This issue needs to be addressed before someone takes advantage of it. Sungodtemple (talk) 13:43, 15 July 2021 (UTC)
I agree that some sort of more formal community process will be needed to screen/approve mentors before this is made live for all. Best, Barkeep49 (talk) 18:09, 15 July 2021 (UTC)
Having a selection process, coupled with existing best practices, could be considered; it is up to you. However, don't forget that selection processes could afraid some people who would be nice mentors, but who are too shy to apply. On all wikis where the mentoring is deployed, anyone is are free to add their names to the list. Some experienced mentors watch the page and check who subscribes. But it is easier at these wikis though, since the list is smaller. Trizek (WMF) (talk) 14:39, 19 July 2021 (UTC)
@Sungodtemple -- we've also thought about the possibilities for abuse of the "claim mentee" feature. We've thought that since a user has to be a signed-up mentor to use it, and since the mentor page is protected to extended confirmed and since the claiming of mentees is a publicly logged action, that that would be sufficient visibility to prevent or surface issues of abuse. It sounds like you think a user right would be safer? I would be interested to hear what others think. MMiller (WMF) (talk) 03:40, 23 July 2021 (UTC)
@MMiller (WMF): thanks for the details on this, I didn't actually realize any roll-out had happened so I was somewhat intrigued as to why I was getting mentorship questions, as I signed up but only to test the feature. (also, would you mind including me on future pings you send here?) Anyway, I think scaling to 25% would be reasonable. The questions I got weren't exactly the best but I wouldn't mind answering 10x more of them, they didn't overwhelm my talkpage or anything. Plus, hopefully more people will sign up with this feature. Elli (talk | contribs) 16:00, 15 July 2021 (UTC)
Arbitrary break
Hi @Calliopejen1, @Elli, @Nosebagbear, @Sdkb, @ProcrastinatingReader, @Sungodtemple, @Barkeep49, @CaptainEek, @Panini!, @Bilorv, @Trialpears -- thank you for your comments and for helping think this through. I'm glad you all think this is generally on the right track, with some challenges to sort out. Here's how I'm now thinking about this: the two big pieces here are suggested edits and mentorship. It sounds like it's pretty clear that suggested edits has been going fine in this test, without any particular concerns or red flags, and we feel comfortable scaling that up to 25% without modification. Mentorship is where we have concerns and ideas. Maybe it could be possible to decouple those two elements so that we don't have to scale them both up at the same time, which would give us some time to figure out the best way to handle mentorship, without that holding back suggested edits from getting to newcomers. We could definitely turn mentorship off entirely (there are several wikis that have the Growth features without mentorship, either because they have pre-existing mentorship programs that they prefer or because they don't have an active enough community to sign up mentors). Or perhaps we could keep mentorship on for, say, 2% of users so that we can keep collecting a trickle of data, even as we scale up suggested edits. I'll check with the engineers on the Growth team to see if that's possible. What do you all think of this idea of decoupling?
Regarding how to communicate about increasing to 25%, many of you recommended bringing that to Village Pump or other big forums. Would any of you be willing to get that discussion going (once we decide here what we're doing)? Or is it better for me to post and get that started?
About some of your other points and questions about mentorship:
While we have lots of good ideas here about how to make mentorship more scalable, I need to caveat them by saying that the Growth team won't necessarily be able to implement them immediately. Most of the Growth team energy right now is going toward suggested edits (a couple exciting projects like this one and this one that I definitely want to talk about in this group once we've tackled the discussion we're in now).
The most important change coming to mentorship in the future is our work on the mentor dashboard. This will be a special page where mentors can see their list of mentees, how those mentees are doing in terms of edits and reverts, and take actions as a mentor (like marking yourself as "away"). I hope some of you can check out the project page and add any thoughts or ideas on the talk page. I've included a mockup of what this might eventually look like.Mockup of mentor dashboard
It sounds like a lot of people like the idea of letting mentors sign up for more or fewer questions. I'll bring it up with the Growth team so we can think about how we might implement something like that.
Several of you noticed that many unproductive mentorship questions come in -- people just saying "hello" or asking how promote themselves. It's definitely discouraging to get those, and it's also discouraging when an answer to a productive question seems to remain unread or not acted upon. I think there's a tough trade-off in there, which Bilorv touched on -- that maybe we have to receive many unproductive questions so that we receive the smaller number of very productive questions that can grow into long-term valuable contributors. I'm interested to hear what you all think about that trade-off.
In that vein is the idea of restricting mentorship to users who have some constructive edits already. While that would cut down on the unproductive questions, there's the risk that some potentially great contributors never get their start because they didn't have a way to ask an important question before trying to edit. I think these trade-offs are something the community here should keep discussing and thinking about.
Regarding newcomers who ask good questions, get responses, but are never heard from again: we hypothesize that many of these newcomers ask their question and leave the wiki, then never find out they got a response because they might not have created their account with an email address, and therefore wouldn't receive a notification off-wiki. This has caused us to think about a couple ideas:
Encouraging users to sign up with, and confirm, their email address. We would do this by updating the account creation interface to encourage email and updating the "confirm your email" message to be clearer. We would definitely still want to make it clear that email is not required, but is useful for wiki communications.
Faster responses through live chat. We think a lot of newcomers are expecting to be in a realtime, or near-realtime chat with a mentor, because they experience live chat help on many other sites (phone company, banks, etc.) This is a pie-in-the-sky idea that we could maybe dig into one day -- something that could be more user-friendly than IRC.
One thing that would likely lead to faster responses is if questions were sent to a centralized place like Teahouse or Help Desk (as CaptainEek says). There is definitely a tension there between "speed to response" and "personal connection". We've moved more toward one-on-one mentorship in the Growth team's work because we think that having that personal connection to another user can make a newcomer feel more comfortable. We've also seen that when given the choice of "help desk" or "mentor", users chose to ask their question to a mentor almost every time.
Indeed! When I first joined some six years ago, I was very impressed by a suggestion to enter a mentorship-like relationship with another experienced user, though it went by a different name then. I ended up not editing much back then, but when I started actively editing about 2 years ago, I looked up the editor and contacted them with a confidence that asking them for help would not be a nuisance to them. I think I was lucky in who I got as a mentor, so we do need a very unpleasant but necessary screening out of over-enthusiastic newbies and rude/impatient experienced editors. Help fora do not have the same effect imo, as they are the default expectation.Even if you/(we?) are not ready to take this to a central forum yet, I think we are missing input from a lot of editors who I would expect to find here. Perhaps send a mass message to editors who have signed up to host at the Teahouse, on Wikipedia:Adopt-a-user/Adoptee's Area/Adopters and had volunteered for the Co-op program? Regards! Usedtobecool☎️04:41, 23 July 2021 (UTC)
Thanks for your analysis here, and considering everyone's feedback. I wouldn't be in favor of a huge decoupling, but maybe 5% suggested edits only, 5% suggested edits and a mentor would be useful to see how much mentors actually help. Elli (talk | contribs) 04:12, 23 July 2021 (UTC)
I would be more than happy to bring this to the village pump, as I think the community would appreciate it more coming from an editor than the WMF. Although it would certainly help for you to be there and actively answering questions :) CaptainEekEdits Ho Cap'n!⚓05:26, 23 July 2021 (UTC)
Decoupling sounds fine to me.The phenomenon of what you politely characterize as "unproductive questions" has come up at the Teahouse too. Restricting the feature to only editors who've already made productive contributions seems like it would defeat most of its purpose, though. There could perhaps be better ways to filter out these questions, though.I assumed that encouraging editors to provide their email was already something we did; if not, that definitely sounds like a step we ought to take. Most sites these days require email on signup, so it's something people are used to giving out.In order for live chat to work, it'd need a lot of editors present. The best existing place for that is probably the #General channel on the Wikipedia Discord. {{u|Sdkb}}talk06:17, 23 July 2021 (UTC)
I think mentorship is definitely the way forwards rather than help desks. As far as I'm seeing, you could turn the dial up to 100% on the suggested edits part of the panel, if a discussion at the village pump is favourable—it's not radically different to suggested edits features we've had in the past and I'm not seeing much objection to it. I think you don't want to try to solve every single issue people have raised with mentorship, particularly as different people will have fundamentally opposing views on what lots of parts of it should look like, but before scaling up to 25% you do want to notify a larger pool of editors so we can increase our mentor list. That hypothetical mentor dashboard looks great, but I don't think we need that or any other improvements before it becomes worthwhile (a net positive) to increase the rollout. — Bilorv (talk) 10:20, 23 July 2021 (UTC)
I think you should make the post @MMiller (WMF) simply because it would be clear, from this conversation, we were acting on your behalf anyway so the advantage of an enwiki "name" would be blunted. I also think scaling up mentorship is an important aspect of a larger test. We know we can handle 2% and to learn anything I would suggest testing some bigger number than 2% get mentors. I'm not opposed to 25% for both suggested edits and mentors as a test but if we're going to limit the mentorship % would suggest 10-15% for this next phase of testing. Best, Barkeep49 (talk) 14:38, 23 July 2021 (UTC)
While I understand that a "hello" message may seem extraneous for some, I think making connections is what helps draw new editors into the community, thus enticing them to engage and participate further. Personally, I think it would be good to encourage laying a bit of ground work between the mentor and new editor. Perhaps the editors can be encouraged to provide a bit of information regarding their interests, so the mentor can help them find relevant active communities on Wikipedia. isaacl (talk) 14:47, 23 July 2021 (UTC)
Isaacl, I agree in theory, but I responded positively to all of my hellos trying to solicit more questions/communication, and none of them responded. :-/ Maybe we could try to track this in the next rollout phase and if none of the "hellos" amount to anything, then we can not worry about them. Calliopejen1 (talk) 16:24, 23 July 2021 (UTC)
My instinct is that even in the best case, the response rate will be fairly low. I'm not saying we should make an ice-breaker type of engagement mandatory. I just think we should establish a welcoming environment for these types of interaction, and not narrowly focus on questions asked and answered. Part of the role as I see it is to help new editors figure out what questions they want to ask. If they don't choose to follow up, that's OK too; everyone can plan their own journey. isaacl (talk) 01:00, 24 July 2021 (UTC)
Thank you all for keeping the conversation going. I'm going to bring all these ideas to the Growth team's meeting on Wednesday (in two days), and I'll be back with some thoughts about what's possible soon after that. -- MMiller (WMF) (talk) 00:42, 27 July 2021 (UTC)
De-couple: first, we make the change to the code so that we can give mentorship to only a subset of newcomers who get the Growth features. That way, we can give the newcomer homepage, suggested edits, the help panel, and the impact module to larger numbers of newcomers, all without being constrained by the mentorship capacity. For instance, we could bump up the Growth features to 25%, but only increase the portion getting mentorship to 5%. This gives us flexibility to figure out how to do mentorship right, without that work blocking newcomers from benefitting from suggested edits. We would hurry to do this in the coming weeks so that we can then start the broader conversation about increasing to 25%.
After that, we could work on these ideas, which I think the Growth team could do within the next six months:
Volume control: allow mentors to sign up for different volumes of mentees. We could allow mentors to choose from "Low" (1x the mentees), "Medium" (2x the mentees), and "High" (3x the mentees).
Opt-in/out: allow newcomers to opt-in and opt-out for mentorship. With this idea, newcomers would first see on their homepage the option to "request a mentor", which if selected, might immediately assign them a mentor. This slight barrier might help cut down on the number of users who are just saying "hello", while keeping the ones who intend to ask more substantial questions. And then newcomers would also be able to "opt-out" to have the mentor disappear off their homepage.
Edit tag on replies: apply an edit tag to mentor replies so that the replies can be filtered in Recent Changes. There is already an edit tag applied to the questions from newcomers, but if we apply a tag to the replies, then that might make it easier for mentors to keep an eye on each others' work. This idea is a bit of a taller order, because it requires parsing talk pages. We still have to think about a good way to do this.
Next, we would experiment with the above changes, talk it over on the wiki, and decide how to proceed with mentorship.
Agree with the first two ideas and the de-coupling, not sure about a tag for replies. Might be able to do it based off of parsing the edit summary? Should work in most cases. Elli (talk | contribs) 03:50, 30 July 2021 (UTC)
I don't have an issue with opt-in mentorship, but I don't agree that we should set a goal of reducing the number of users just saying "hello". I think mentors should be ready for many of their responses not to garner any follow-up. I know it's discouraging, yet I feel it's a practical reality. isaacl (talk) 04:00, 30 July 2021 (UTC)
Hi Miller - that all sounds fine (if anything, we'll again need to cap Mentors if we only need 150% more for the mean time). The Community can handle a psuedo-perm on our side for mentors without any immediate action needed by the Growth team - we need to actually agree what we're going to have, which will require a Community discussion, and so best not to wait for that before scaling, but it doesn't require any technical changes. I don't think a reply tag is needed - you could no doubt garner some benefit from it, but in terms of opportunity cost, it seems like it could be steep. Nosebagbear (talk) 10:54, 30 July 2021 (UTC)
I think volume control is quite important and the proposed system sounds good to me. For decoupling I think that to sounds sensible. I don't see any reason not to increase the growth tools without the mentorship part to everyone though, except for AB testing. The opt in/opt out option sounds like about the right level of obstacle in my very non-professional opinion. I would consider tags on reply as lowest priority but potentially useful. --Trialpears (talk) 15:33, 1 August 2021 (UTC)
I think decoupling makes senes and I would love for us to move to the next stage because I am hopeful there will be community consensus and I think this change a positive one that I'd like to see benefit more new users. Best, Barkeep49 (talk) 21:50, 9 August 2021 (UTC)
@MMiller (WMF):, I think we weren't really sure of the timing. I'll go to WP:Teahouse and WP:Helpdesk and ask for another 150% mentors - that will, again, not be a problem. I think we'd be fine to move to 5% mentors and 25% platform in any time from say 2 weeks time (obviously if decoupling code on your side takes longer it takes longer!) Anyone here think that's problematic? Nosebagbear (talk) 08:41, 13 August 2021 (UTC)
The next step is to bring this up at the appropriate Village Pump. I'm less experienced with the Village Pump process than many of you, so I have some questions about the best way to go about:
Which Village Pump? The original 2% test was announced here at VPR by User:Nosebagbear -- is that the right one?
Who should post it? I am able to do it, but perhaps one of the volunteers would know better how to frame our announcement.
Is this more of announcement or more of a discussion? Will we be waiting to achieve some kind of consensus, or do you expect most people to have no concerns?
Meanwhile, we'll also plan to work on the improvements to mentorship mentioned above, that will help us roll that part out to more newcomers. This work will likely take several months to unfold, before we deploy it and then start feeling comfortable increasing the mentorship percentage on this wiki:
Mentor dashboard: allows mentors to see their list of mentees and customize their experience.
Volume control: allow mentors to sign up for different volumes of mentees.
Opt-in/opt-out: allow mentees to opt-in or -out of mentorship.
VPR is best. A number of the people you pinged have a good history with successful major proposals; one of them can probably advise better on how to word it. I can't imagine anyone having much of an objection, but something that reads like a discussion is better than something that reads like an announcement, for optics if nothing else. I can't personally imagine an RfC or something is required unless there's actually a dispute, though. ProcrastinatingReader (talk) 00:58, 17 August 2021 (UTC)
Congrats on that work @MMiller (WMF). Briefly, VPR seems right to me, I think you should be the one to post it as, I see this as more of a discussion. I think it's the community's first opportunity to widely know about this, to ask questions (many of which you no doubt have answers for or one of us who've been following closely might have answers) , and otherwise wrap our collective heads around the concept. I expect there to be a fairly positive reaction because the initial data is good and this ask is a reasonable progression. But it's also an important progression as I would expect the next ask to be some level of full adoption. Best, Barkeep49 (talk) 00:59, 17 August 2021 (UTC)
@Barkeep49: - I agree with your comments that MMiller should be the one to raise it, but beyond that, I think he (on behalf of the team) deserves the broader awareness of not merely a great project but also an example of how it's done. That latter aspect includes how WMF staff can take a project through community review directly Nosebagbear (talk) 18:38, 20 August 2021 (UTC)
I agree this has, so far, been a positive collaboration from my perspective as a community member. Hopefully the team on the foundation side feels similarly. Best, Barkeep49 (talk) 19:32, 20 August 2021 (UTC)
Yep, I concur with Barkeep and ProcReader—you opening a VPR discussion sounds good. For framing, I'd say just make sure the central question (are we ready for adopting at 5%/25%?) is clear at the top, give a quick overview of the features that doesn't assume any prior knowledge and the recent results, and provide links to here for anyone who wants to dig deeper. {{u|Sdkb}}talk03:05, 17 August 2021 (UTC)
If you like, post a draft in a new section on this page and people will offer thoughts. The initial wording is important. Frame it as a proposal saying that the Growth Team has conducted a trial of [brief description with link] and propose a change to 5%/25% (with brief explanation). Johnuniq (talk) 03:25, 17 August 2021 (UTC)
I've had 2 prior experiences with the Village Pump (from what I remember) about small proposal ideas. With these two experiences, I concluded that the usuals there ask a lot of questions and usually stray from making changes when possible. Not sure if that's incorrect and I lucked out, but I agree with Jognuniq for this reason; a draft we can work on to make a proposal that's explanatory as possible so other users don't have issues with it. Panini!🥪11:31, 17 August 2021 (UTC)
Hi all -- I started drafting up a proposal like @Johnuniq recommended. There are still a few things I want to add (even though I think it's probably too long), but I would be grateful for any suggestions, or if you just want to make edits directly. Feel free to post here or at the bottom of that subpage. Thank you! MMiller (WMF) (talk) 04:36, 26 August 2021 (UTC)
@MMiller (WMF): It can either be included as a single line towards the bottom of the VPR request, or, on-passing, a dedicated request at WP:AN - but we'll probably want another 25 or so mentors. Nosebagbear (talk) 10:46, 22 August 2021 (UTC)
@MMiller (WMF): looks good - only potential improvement that jumps out to me would be after with a revert rate of about 6% put in brackets what the usual revert rate is for editors in this category (something like 22% I believe? I remember it nicely rammed the point home) Nosebagbear (talk) 00:48, 1 September 2021 (UTC)
Edit suggestions settings not completely saved immediately?
I just tried filling out the form to get "Suggested edits". I selected three topics of interest and checked all the boxes for Easy and Medium-difficulty editing. That seemed to work and edit suggestions showed up, but I then (without refreshing the page) clicked the "3 topics" button, opening a dialog to see or change my selected topics. In this dialog, no topics appeared selected, and, when I closed the dialog without making any changes (clicking the "X" button, not the "Done" button), the "3 topics" button no longer said "3 topics" but rather "Select interests", indicating that the selected topics had been forgotten. I then clicked the "Easy, Medium edits" button just to the right of "3 topics"/"Select interests" and my edit difficulty settings were likewise forgotten. This seems a small bug, but one that might be discouraging for a real newcomer.
I then used both buttons again to set the topics and difficulty settings, and that seems to have worked. I don't think I can reproduce the bug, without creating a new account.
Hi @2d37 -- thank you for checking out the Growth features. I'll ask the engineers on the Growth team what they think of your experience, and see whether we can reproduce it. In the meantime, I would love to hear any of your thoughts on the rest of the features. Let me know! MMiller (WMF) (talk) 01:24, 13 October 2021 (UTC)
@2d37 -- we just tested this ourselves and we were able to reproduce it. It seems like it only happens for a new user the first time they select topics (which is the most important time, of course). Thank you so much for the report -- we'll prioritize fixing it soon. Here's the task where we filed it, if you want to follow along: https://phabricator.wikimedia.org/T293163MMiller (WMF) (talk) 01:47, 13 October 2021 (UTC)
Follow-up questions in the same thread?
Something that I've seen happening a few times with mentoring is that a new user will ask a question, I'll reply, and then they'll reply with a follow-up or with thanks and that'll create a new level-2 section header. Could the software be changed so that replies automatically go in the same thread, or at least in a new level-3 section within it? {{u|Sdkb}}talk19:02, 18 October 2021 (UTC)
@Sdkb, I think the best improvement will be made when the ReplyTool will be available for everyone. Then it will be easier for newcomers who come back to your page to find where to reply. At the moment, the only way to edit a talk page is to understand that you have to click on [edit source] on the section title. This may explain why people go back to their Homepage and use the form there to contact their mentor. :) Trizek (WMF) (talk) 14:38, 20 October 2021 (UTC)
Thoughts about the newcomer homepage
I viewed the newcomer homepage today and was very fond of it. Everything is easily laid out and easy to follow along with, and a mentor was readily available; I believe it'll be very helpful for newcomers and is definitely a step up from the Wikipedia:Community portal, which is very cluttered and confusing.
The page really puts Wikipedia in a good light, making it as friendly as possible and shows Wikipedia's true purpose. Info like "Your impact" and "Help make Wikipedia better" would definetely make a newcomer feel good, even if it was typo fixes.
One note, though, and I'm unsure on whether or not it's covered; the get help with editing section is resourceful on teaching about formatting, style, and the expansion of articles. I believe simple guideline pages should be just as readily available as this one. Often, I'll see a user on the Teahouse that has made a fleshed-out draft about something or someone that is not notable enough for inclusion. It's something that needs to be caught before newcomers learn the hard way. The page should show info about WP:NPOV, WP:N, WP:CON, WP:VD, and WP:NOT. When I began contributing, I was often working hard and then had my edits abruptly removed because I didnt now about WP:NFC or similar. Although there's a search bar, some might not search either because they don't know what sepcifically there looking for or they assume some silly rule doesn't exist in the first place. Panini!🥪14:20, 14 May 2021 (UTC)
Definitely. We lose so many editors and waste so much labour time (on both reviewer and writer ends) by people not having a clue what our scope is, what our quality standards are and where we actually want help (usually by improving an existing article, not writing another one based on a defective cookie cutter). Sometimes I think that us long-term editors are just the ones who luckily stumbled across something where their first attempts at hard work just so happened to be useful (for me, notable episodes of The Big Bang Theory). — Bilorv (talk) 15:06, 14 May 2021 (UTC)
@Panini! @Bilorv -- thanks for trying out the features! It helps to hear which parts you think will be the most valuable. I think a challenge you're touching on is that there is so much we want to tell newcomers, but they have limited bandwidth and patience to absorb it all. We've heard from newcomers that pages with lots and lots of help links can be overwhelming. That's why we limited the homepage to just having the five most important links. The list we have now are the five that User:Sdkb recommended, but we could talk about if other should be swapped out. MMiller (WMF) (talk) 21:24, 14 May 2021 (UTC)
Absolutely—I hate all of these welcome talk page templates and policy/guideline directories that go on for miles. Five sounds good. I think human contact early on is important too. If you read WP:N top to bottom one thousand times then, without actually seeing it in practice, you will still not know where the community sets the notability line (i.e. is two sources for this topic enough? four? eight?). But if you read enough to get an idea that "notability" matters, realise that you need to look for sources first (not write an unreferenced essay and submit it to AFC) and ask someone "is something like [these links] enough for notability?" then you will get your answer. (Or, ideally, you do none of this and instead start with an easier task of improving an existing article...)
So some overview of the key information in as much detail as someone will actually read, and then early contact with a mentor before someone has invested hours of their time into an idea that just will not pass muster. — Bilorv (talk) 22:52, 14 May 2021 (UTC)
Bilorv, you are right about me! I started with copy-editing articles. Anyways, back to the topic. Which policy pages should be included? I'm thinking: WP:NPOV, WP:Notability, WP:MOS, WP:AGF, and WP:NOT. These are the main issues I see newcomers run into:
NPOV is something newcomers often do not realize applies to articles. Many newcomers do COI editing, not realizing that one must stay neutral.
Notability is, in my opinion, an obvious pick: We've all seen those who, again, write about themselves or some YouTuber they like, not realizing the lack of sources.
AGF should be included as it allows newcomers to see the 'community' part of Wikipedia, one that most never realize exists. It also helps avoid shouting and attacks.
I think it would be difficult to compact every guideline and policy into one section (on my Chromebook there is an empty rectangle of space for putting something like this, but it's definitely not the same elsewhere). A quick little thought I had to make it accessible: I would propose a section similar to the "Get help with editing" section, with 5 core policies being listed; in that section, maybe a "Rule of the day" portion that gives the user info about new content to learn daily, being able to learn more and more as they progress without the need of shoving it all in their face at once. Maybe something much better than this (you all are in charge here), but I believe its important, as mentioned above, that rules are given to newcomers in a casually paced matter. Panini!🥪03:25, 15 May 2021 (UTC)
My top picks for this would be these five (in no particular order):
Keep it simple ...."Wikipedia does not publish original research. An encyclopedia is, by its nature, a tertiary source that provides a survey of information already the subject of publication in the wider world. Accordingly, and because Wikipedia is open to editing by anyone, we require that information be verifiable in reliable sources. Ideally, all information should be cited to reliable sources and to evidence that is verifiable. However, we only mandate citations for quotations, material whose verifiability has been challenged or is likely to be challenged and contentious material, whether negative, positive, or neutral, about living persons. Sourcing requirements are significantly stricter in such articles on living persons. Please note though that we strongly discourage people from writing about themselves, their friends, bands, websites, companies or organizations, pet projects, personal vendettas, campaigns and any other topic in which they have a conflict of interest." --Moxy-
Hi Panini!, Sungodtemple, Bilorv, and Moxy -- thanks for thinking all this through and trying to prioritize which policies are more important to teach. Because we have those five slots on the homepage (and in the help panel), we think that some of the links need to be more practical than policy pages, e.g. information on how to even click that edit button and start using the visual editor. We want newcomers to be able to plunge in and edit quickly so they can get excited about being involved. Sdkb went through and recommended the five links you'll see in the homepage if you visit it now:
I think this is a good list, but if you all think this should be different, please do continue the discussion here. Also, Panini!, I like your idea of rotating/refreshing learning content so that people can learn something new every time they visit the homepage. I added your idea as a comment on this task for us to think about in the future. -- MMiller (WMF) (talk) 01:17, 25 May 2021 (UTC)
Thanks for the ping, MMiller. To clarify just a bit, I'm not committed behind those links as the best possible ones; I just recommended the switches from WP:MOS to the beginner version and the old article wizard to the current one as improvements within the existing framework. And I still feel pretty iffy about the ordering—style is really the last thing we ought to be teaching newcomers, not the first, since it's the easiest thing for others to correct and probably the area where a newcomer who does it wrong is least likely to get bitten.
One other thing to note is that all the links currently point to VisualEditor guidance. I don't personally mind the nudge toward that as the default (others might), but if a newcomer changes their preferences to start editing in Wiki Markup instead, the panel should at least have the functionality to adapt and start offering the equivalent Wiki Markup help pages instead.
Very quickly skimming some of the discussion above, I'm seeing a lot of links to policy/guideline pages. While we do at least try to keep those accessible and beginner-friendly, we inevitably fail, since there's a tradeoff between beginner friendliness and comprehensiveness, and the latter always wins (because WP:CREEP). So I'd recommend instead using Help:Introduction to policies and guidelines/1 as the entry point, since it offers concise summaries of the most essential ones. {{u|Sdkb}}talk02:15, 25 May 2021 (UTC)
'How to create a new article' seems iffy. Users have to be autoconfirmed in order to do so, and according to the image on the project page, there is a 20% chance of a registered user making an edit. Accounting for editors like vandals and COI editors, I'd expect less than 10% will actually become autoconfirmed in the first place, not to mention they are likely to use AfC instead of creating an article directly in mainspace. And assuming they actually want to create a new article. Creating a new article is actually very similar to expanding one anyways, so I think that one should be removed, or at least held off until the user makes a few edits or becomes autoconfirmed. Although, this may lead to that WP:VESTED feel. Sungodtemple (talk) 02:32, 25 May 2021 (UTC)
I would recommend normal pages edited by the comunity with normal formats that work for all in mobile view and desktop view. Lets try to encourage academic readers to start writing. Link them to real Wiki style pages with zero learning curve as we should steer clear of VE only pages...link our normal pages that cover both like Help:Editing over Help:Introduction_to_editing_with_VisualEditor/1 that requires loading many pages for data of only one editor mode. "VisualEditor's effect on newly registered editors/May 2015 study". Wikimedia.org. Wikimedia Foundation. Retrieved 2 August 2016.Moxy-10:12, 25 May 2021 (UTC)
I don't care about it being VisualEditor pages, but I do disagree with "How to create a new article" for the reason Sungodtemple gives—the answer is "don't. You wouldn't drive a bus before you've learned how to drive a car. Edit normally for a while before you think about creating new pages." I also disagree with "How to add an image" because the answer is "your image is a copyright violation. Do not upload it" in 99% of cases. If there is any page linked at all, it needs to be extremely clear how copyright works and why "I took a screenshot of this video game so the image is mine" and similar misconceptions are not true. Otherwise you're just encouraging copyright violations on Commons, which are very slowly dealt with and very inconvenient for the en.wiki community (as few of us are admins on Commons). — Bilorv (talk) 11:10, 25 May 2021 (UTC)
Thanks for continuing this conversation. Since there are a bunch of different opinions here, I think we should wait to make changes until there is some more consensus. Then we can easily update the links. We can put any five links with any labels in any order, so I think the easiest way to express the desired list is like I did above:
In the future, we plan to make available a form so that administrators can modify this without the Growth team being in the middle (that project is here). But for now, we'll need to make the change on our end. Please let me know! MMiller (WMF) (talk) 18:33, 26 May 2021 (UTC)
Bilorv, re image copyrights, looking through the tutorial, the Wiki Markup version covers at least the basics, but the VE version shoves it down to a small-text "important caveat" notice. The VE version should probably be edited to be more like the Markup version (while still remaining as short/concise as possible). I just checked and uploading an image with VE does require affirmatively checking a "this is my own work" box, which is at least something. {{u|Sdkb}}talk19:03, 26 May 2021 (UTC)
From reviewing at AfC and invariably finding that any non-free image is uploaded to Commons under the claim of "this is my own work", I feel no confidence that the VE notice would have any effect at all. The "Important caveat" in small should be moved to large bold at the top, with the notice "Before reading this:". — Bilorv (talk) 19:07, 26 May 2021 (UTC)
How much do you think the issue is people mistakenly thinking a work is actually theirs vs. not caring about copyright and lying vs. just trying to click through and not reading? {{u|Sdkb}}talk19:57, 26 May 2021 (UTC)
No-one's lying intentionally, but they are both clicking through with little attention paid and choosing not to think about it, "I don't care; just let me upload the image; why are there so many steps?". I never cease to be astonished by how little people know about copyright. I guess it comes from an internet in which copyright law as it applies to YouTube copyvios or people copying full paywalled news articles into internet fora is never discussed. Then again, if they don't know what Commons is then even a full understanding of fair use won't help because the images usually are legal in a Wikipedia article under U.S. fair use (so far as I can tell) but fail WP:NFCCP. — Bilorv (talk) 20:07, 26 May 2021 (UTC)
A side note from my previous comments, I'm not proposing to replace Sdkb's list; I see those five just as absolutely important and I believe they're the cause for the productive and constructive edits in the statistical results. I was reporting that there should be a list offset from this one that clarifies that rules and policies exist and what to keep in mind. However, if there isn't a way to introduce these in an efficient and understandable way it's not the end of the world. Panini!🥪19:26, 26 May 2021 (UTC)
Updating the help links box
Can we return to this thread for a moment? asilvering has highlighted issues with the help link box and I feel it's within relation to this section. While the help link box gives a good understanding of how to edit an article, it doesn't explain how to do it properly. They're taught how to add an image but not where an image is necessary; how to add a citation but not what's reliable and notable; how to create a page, yet everybody at the Teahouse will tell you not to as a newcomer. This little sidebox should be a very helpful tool to the newcomer, but it can be much more helpful. Being told something you did was wrong can be disheartening and frustrating. How can we make it better by having the core rules to understand when additions apply the best right at a new user's fingertips? Panini!🥪13:52, 9 November 2021 (UTC)
Feedback from a newbie
Hi Growth Team! I'm new here and I'm apparently one of the 25% of new-user guinea pigs with the new homepage. I have some feedback to share, and I hope this is the right place to do so. (Though I only see what looks like experienced editors on this talk page?) Overall, if the alternative to this homepage is "nothing at all," this is clearly superior. But I am already finding it frustrating in a way that simply being abandoned to my own devices would not have been.
I think the "suggested edits" widget is a great idea, and I want to love it, but I really think it needs some rethinking. It appears to be completely auto-generated from various "this article needs help" categories - but this means that new users are immediately faced with some of Wikipedia's worst articles, the ones in need of the most help. We're told to make some copy edits, which are ballparked as a 5-10 minute task, but the articles that are linked are in need of so, so much more work than that. One that keeps coming up in mine right now is Ichabod Crane. This does not need minor grammar fixes. It needs a complete overhaul. If there are articles in this pile that could be meaningfully improved in under ten minutes, I have yet to be offered one. I'm sure the aim here is to communicate to new editors something like "you can make a useful difference right away!" and "even a few minutes' work is worth doing," but it's extremely dispiriting to be handed a Start-class article out of the gate. I could fix some grammar errors, but what would be the point, when an editor is going to have to rewrite the section entirely to bump the article up to C- or B-class? It makes editing Wikipedia look like a hopeless slog.
I think the first several suggestions this widget gives should be curated or filtered in some way - perhaps limiting results to articles that are B-class? These articles are pretty likely to still contain a grammar error or a link that could be improved or something similarly doable in 5-10 minutes, and they serve an added bonus of serving as a better model for new editors. (It's much easier to learn standards and conventions when you're started out on articles that follow them more often!) I would also suggest renaming this first set of easy edits from "copy edits" to "minor edits," because it seems to me that the purpose here is to get new users to start small and build confidence. Since the meaning of "copy editing" on WP appears to include "significant rewrites for clarity and length," and the Category tags that are populating this widget include a lot of requests for these more substantial edits, it seems to me that this should be a second newbie-onboarding step, not the first one - something the widget works up to after suggesting a handful of the easier edits.
Please, please give us a link to a Manual of Style in the Help links box on this page. I don't know why "how to write a good article" and "how to create an article" are here - I have been given the strong impression that it is not desirable that I create a new article, as a new user, but instead work on editing already extant articles first - but not a link that would be helpful for editors doing the "easy" copy edit tasks. I see someone above on this page saying that WP:MOS is long and overwhelming. I agree! A short, simple, accessible guide, with links to the full one, would be most useful. But even a link to the long and overwhelming guide is better than hoping new users flail in the right direction and manage to find something that can answer their copy editing questions. I didn't even know such a document existed, let alone how to find it, and was having to reverse-engineer what the MOS might be from the articles featured on the main page. asilvering (talk) 22:59, 8 November 2021 (UTC)
Not a WMF staff member, just an editor, but I'm really glad to see these comments Asilvering and I'm sure they'll be useful. I do like the class idea, though I'd maybe go with C-class as B-class articles might not really have that many grammar issues. In any class you will find really varying levels of easy corrections to make, along with many articles whose class ratings are out of date, so things can be really randomly hit-and-miss (our six-million strong set of articles is categorised in lots of ways, mostly in quite crude ways because of the size and scope, and most designed around more experienced editors). The disadvantage to being given better-quality articles is that you will undoubtedly find five or six things that look like natural improvements to you (say, adding more detail to a plot summary of a TV show) and they will all be reverted as violating very specific rules you couldn't know yet (MOS:TVPLOT sets word limits in the fourth paragraph). Short of heavy investment in some tool where experienced editors can specifically flag up "this is a task I'd like a new editor to spend 10 minutes fixing, here are some instructions", there will always be a lot of misfires on automated suggestions. I agree with your suggested link changes (and you're not the first to make this argument). — Bilorv (talk) 00:41, 9 November 2021 (UTC)
This is really helpful; thanks, Asilvering! I think the point about the suggested articles being ones in really bad shape is something that's definitely worth heeding. In addition to the issue you identified of it making it harder to learn the MoS, some other drawbacks are that really obscure articles tend to be on less interesting topics, newcomers who make errors there are less likely to have an experienced editor notice and help, and edits there will have less impact (even setting aside the impact module, I think most people can intuit that they aren't often read).
One thing the Growth Team may want to consider is providing newcomers with the information that tags are community-added, and sometimes they can be wrong. If you come across a tag-bombed article that doesn't actually seem like its main problem is underlinking, it's fine to skip it. Cheers, {{u|Sdkb}}talk03:24, 9 November 2021 (UTC)
Hello @Asilvering! Thank you for joining the community and for taking the time to write down your thoughts as a newcomer. I'm Marshall Miller; I'm the product manager for the Wikimedia Foundation product team that build the experience you're talking about, and I've been working with all the volunteers on this page about how to implement that experience on English Wikipedia.
May I ask how you found your way to this discussion? It's always so valuable to hear from actual newcomers, and I'm wondering how to encourage more people to provide their feedback on the features. I would also be interested to hear anything else about your journey, like why you decided to start editing and any prior experience you had with Wikipedia.
You're right that the objective of "suggested edits" is to help newcomers start small, see that they can make a difference, and build confidence. I haven't yet heard the perspective you're bringing: that it can be dispiriting to be confronted with an article that needs an overwhelming amount of work -- and that you might be able to tell that a necessary reorganization of an article could wash out any small improvements. Our thinking was that an article that needs a lot of work can make it easier for a newcomer to find something in the article to do. One challenge has been to make sure newcomers know that they are not expected to complete the copyediting of the article; but rather that any small improvements are worthwhile. Thanks for bringing up the idea of a more a narrowly curated set of articles that are in better overall condition with a couple of obvious errors to fix. Community members here are actually equipped to implement something like that: administrators on this wiki are able to control which templates are used to populate the suggested edits feed, and I could imagine them developing a specific template to put articles in that feed, and then configuring the feed to show only those highly curated articles. And perhaps that template could contain specific instructions about which errors the article contains (e.g. "This article misspells the word 'conceive'.") What do you think of this idea?
I also wanted to ask: did you notice/use the "help panel" in the bottom right of your screen while attempting suggested edits? It is meant to surface some quick tips about how to complete a copyedit.
Another way that we're addressing this challenge is through "structured tasks". These are specific suggestions for articles generated by algorithms, meant to make it quite clear where a newcomer's attention is needed. One that we're piloting is called "add a link", which suggests phrases to turn into wikilinks. Another that we'll soon start piloting is called "add an image", which suggests an image to add to an unillustrated article. We're testing these in other languages before bringing them to English Wikipedia. I would love to know what you think of these ideas. Would they help someone who is totally new?
I understand your question about why we include guidance for how to create new articles, if we don't think newcomers should start by creating them. The reason is that whether we think it's a good path for them or not, a very large portion of newcomers are determined to create a new article, and so we think it would be better for them to see the guidance than to attempt without any guidance. Do you think that's the right approach? And with the manual of style, you can see in this conversation that we did indeed discuss which links to include. In addition to @Bilorv and @Sdkb, @Panini!, @Moxy, and @Trialpears participated in that conversation (pinging them in case they want to check out your perspective and change the links).
How I found this: I can't quite remember how I googled my way here - "wikipedia homepage widget"? Not sure. I started looking in the first place because of the widget issues I've mentioned.
How to encourage feedback: honestly, I wasn't sure anyone would want individual user feedback because of how there didn't appear to be any way to give it. It felt like perhaps the ability to respond was being intentionally withheld! I'm sure you'd get more if you made a feedback talk page for the "suggested edits" widget and put a hyperlink to it at the bottom, right under the "Other users have noted these articles need work" bit, or wherever else makes sense. I suggest a specific, new talk page for this because I think it's important that the page be mostly new-user feedback and responses rather than other planning stuff, experienced editors talking to each other, and so on - whatever you can do to make new and easily confused people think "this is the right place" and "I'm not interrupting anything important or bothering anyone if I post here."
Why I started: I've been using Wikipedia for ages and I've occasionally jumped in to fix a typo or a dead link or something. I don't think I ever had any desire to write articles from scratch, but I did occasionally want to make more involved edits, and didn't do so because it felt a bit weird to do that just under my IP, or because I didn't know how to make that specific type of edit and didn't feel like trying to learn at the time. Recently I've been using it more often and noticing stuff that's extraneous, outright wrong, or just really overly wordy, in ways that wouldn't take me much time to fix. But I figured that deletions or much-condensed summaries would just be reverted as vandalism unless they came from an established account with a history of edits that aren't nonsense. In the end maybe that's not quite true - at least in the noob guide stuff, there's a lot of "assume good faith" that, honestly, was counter to my impression of how WP operates.
Suggested edits: sounds great. Things like "this article misspells this word" sound like it would be more effort for community members to tag than it would be for them to just fix, but an "easy edits for newcomers here" tag seems simple enough to use? Or, if you think it would be helpful for them to be more specific, some simple ones like "newcomer task: fix dead links", "newcomer task: standardize spelling", "newcomer task: find a source", and so on seem pretty doable to me.
Help panel: I don't think it's possible to not notice this, at least on the desktop version. It pops up open by default. I didn't find any of these tips helpful, but I think they probably are helpful to some new users who don't have much experience with copy editing. Maybe the last panel would be a good place to put a link to a newbie-friendly manual of style?
Structured tasks: This sounds to me like a tutorial, but better, because you're actually doing real work right away. I think this would help a lot.
Guidance on creating new articles: yeah, I understand this. What I meant was more that I don't see why there are these links to the exclusion of links that would help new editors do the things you actually want them to do. I realized only recently looking at these links again that the MOS is indeed linked in the first article, "How to write a good article." But I didn't click that, because I wasn't trying to write an article, because I'd been told not to write an article.
I do wonder if the concern about the users who want to immediately make an article is worth the effort, though. Do you really want to encourage those users to stick around? I assume you have some data you could look at to determine what makes sense here, but it strikes me that someone who wants to make a completely new article as the first thing they do on WP is very likely to be advertising their self-published novel, performing conflict-of-interest edits about their employer, vandalizing, goofing off, just generally misunderstanding what WP is about, whatever. It might actually be in WP's best interests to show these people the door as soon as possible, before they get too invested in their article, while ensuring as much as possible that they leave with a generally positive opinion of the place. (Because they may come back wiser in the future.) That's not to say that WP has never had really helpful, long-term editors start out by making a new article first. But what's the return on your time and effort, here? And are your metrics to determine this measuring the things you actually want to measure? Sure, a large proportion of new editors want to start by making a new article. But you're never going to convert most new signups into long-term frequent users, on WP or anywhere else. Maybe these new-article-first users are just self-selecting out, and you shouldn't worry too much about them.
My guess would be that the people most likely to stick around usefully and long-term are the ones who start by editing an already extant article or talk page. And I'd be worried more about losing someone who does this in good faith and gets burned, or who shows up full of naive exuberance and immediately gets bogged down in policy or intimidated by the scale of WP, its jargon, and so on, than about losing the person who made an article about their sister or their dog or what-have-you. The person who adds a truthful but unsourced statement, or who thinks they're being helpful when they add irrelevant trivia - I'd be more concerned about these new users, because they're going to get hit with reverts almost immediately and get discouraged. And I think these are the new users that are more likely to stick around, if encouraged to do so. I think the structured tasks will help here a bit, because those users can do a bunch of simple edits that aren't likely to be reversed, so it will feel less like WP is out to get them. Trying to capture people who fix typos without an account or the ones who show up on talk pages with suggestions/questions might also be better effort-to-reward than training the folks who are determined to do exactly what you don't want them to do and write a new article. But again, that's based on my assumptions - hopefully you have data that can help you answer these questions, or, if not, the ability to start recording it. asilvering (talk) 21:17, 11 November 2021 (UTC)
