Talk:Growth/Personalized first day/User profile

Hi @KStoller-WMF! Since you mentioned yesterday that this was a potential project for the coming year, I wanted to stop by and offer some early thoughts.

Overall, this project would touch on a lot of cultural aspects, which introduces some uncertainty as wikiculture can always change. One path I could easily see this taking, though, is that if a structured user profile is offered to newcomers, it'll quickly become associated with them, and no matter how good its features, users will want to stop using it as soon as they become experienced enough to learn its reputation.

Newcomers are often very eager to create user profiles, so I don't think you'd have any issue with uptake. But before we get too excited about that, uptake alone wouldn't be enough. The community has made very clear that userpages are not just a blank canvas, but rather required to aid in the development of the encyclopedia. If an easier way to make userpages resulted in newcomers doing so, but then never going on to make edits to articles, that doesn't help us.

Worse, the ability to create a userpage is currently an incentive for newcomers to learn basic wikitext. If an easier way to create a profile was introduced, this incentive would be removed.

Given all that, I'm not sure this is something to prioritize, given that it's high-effort for uncertain benefit and has drawbacks. I'll note that, while an editor proposed it for the 2022 wishlist, it didn't do all that well — in the bottom half in terms of number of votes, and several opposes despite those being uncommon for wishlist items. For lower-effort alternatives, I think the community could perhaps improve the editnotice for blank userpages to help alleviate the "what do I do with this space?" question; your input would be welcome on that, but there would not be any technical elements that'd require developer support.

Cheers,

Thank you so much for spending the time to attend meetings and offer your thoughts!

the ability to create a userpage is currently an incentive for newcomers to learn basic wikitext. If an easier way to create a profile was introduced, this incentive would be removed.

This is a very important insight. I agree that this is a project that has uncertain benefits, I could even see this having a negative impact if approached in the wrong way.

I would definitely want to try a small experiment before we invested a lot of time into a project like this. Something like allowing just one small part of the userpage to be structured, like a mini guide on how to add language fluency or something like that. Perhaps that would reduce the initial barrier to entry, and then the newcomer will be more likely try more unstructured edits to their user page.

That all being said, I'm not sure even a small experiment will end up getting prioritized next fiscal year. As of now there are other projects that we already have some research and experiment data to suggest they will be successful in improving newcomer retention.

For lower-effort alternatives, I think the community could perhaps improve the editnotice for blank userpages to help alleviate the "what do I do with this space?" question

I think this is a great idea! I haven't spent that much time thinking about this, but I'm happy to spend some time considering this and chatting with our UX designer about improved copy for the edit notice if you think my feedback might be helpful.

I'm happy to spend some time considering this and chatting with our UX designer about improved copy for the edit notice if you think my feedback might be helpful.

Yes, certainly! The current notice text was a little tricky to find. It appears to be at w:MediaWiki:Newarticletext, producing what you'll see at e.g. here. I notice that there's nothing at all talking about what a userpage is for or what you should put on it.

One area for technical improvement that I see: There currently doesn't appear to be any way to present different text based on whether or not it's your userpage vs. someone else's, as I don't think {{subst :REVISIONUSER}} would work in an editnotice. That limits how tailored we can make the message. Is there a Phabricator ticket that'd address that? Cheers,

The Foundation pursued this idea before. The idea initially received a chilly reception from the community. The team built it anyway. On deployment day there was an epic-level manure explosion and the project dropped stone cold dead.

First I'll address the chilly reception. Everyone is familiar with social networks. It defines a large portion of the internet. Wikis kinda resemble social networks. You've got a big community of people.... on the internet. Except our purpose is very different. If a social network is like nightclub, wikis are like an office space. The expectations and culture are very different. We find it disruptive to our work when people show up and behave as if the wiki is a social network. We firmly reject any characterization of the wiki as a social network. On EnWiki we have a policy link dedicated to this: WP:NOTSOCIALNETWORK.

We're not a social network, and we don't want the wiki to resemble a social network. Anything that makes the wiki look more like a social network increases the likelihood that people will treat it like a social network. And as I said, that disrupts our work. We tend to be extremely skeptical when the Foundation initiates any work which is inspired by social networks, or which makes the wiki look more like social networks.

And to be honest, we look down on social networks. We're doing important work to benefit humanity. We can have a snobbish disdain for social networks, and a disdain for anything with a whiff of making the wiki resemble a social network.

Then there's another point. A lot of things about wikis do initially come across as unusual and unfamiliar for new users. However the concepts of a wiki very quickly come together in a very simple way once people start editing. What's a wiki? It's just a bunch of pages. What's a page? It's just a text file with a title.

What's an article-talk page? It's just an article-page with a different tile. When you encounter an article-for-deletion discussion page for the first time, it's not really new. It's just an article-page with a different title.

Once you learn a little bit of editing, a page quickly becomes a very powerful and flexible thing. Anything that works in one place works exactly the same everywhere.

What's a userpage? It's just an article-page with a different tile. It's whatever you edit it to be.

Skydiving is a weird hobby, most people don't enjoy jumping out of airplanes. But if you run a skydiving foundation, that is your target audience. If you run a wiki foundation, your target audience is people who think editing-an-encyclopedia is a fun hobby. Your audience think it makes perfect sense to use a wikipage as a userpage/profile.

And here's why the last project exploded in a spectacular fashion: Staff have a bad habit of thinking it's a neat idea to add a pretty PageImage() whenever the software mentions or links to a page. Are we showing article editing suggestions? Let's toss pretty PageImages on it! Are we showing recently edited articles? Toss pretty PageImages on it!

As editors, we have a very different view. First, we don't toss decorations on links to make them more click-baity. We have no interest in maximizing clicks to maximize advertising revenue. But more importantly, we are acutely aware that our images are WP:NOTCENSORED. When the software blindly retrieves a PageImage(), it may be retrieving an image of an asshole, a vagina, fetish-bondage, a string of semen running across a neckline, anal fisting, an image of Muhammad, Piss-Christ artwork, a maggot-infested wound with exposed bones and tendons, the mangled bodies of warcrime victims, or anything else. Britain even imposed a nation-wide block against one of our PageImages as alleged illegal child porn. See en:Internet Watch Foundation and Wikipedia for encyclopedic coverage of the incident. (Or don't see the article, the image involved is reproduced there.)

When the profile page was deployed it included a list of recently edited articles... with PageImages of course. On deployment day, people who had edited completely random pages to revert vandalism discovered they now had shiny new profile pages.... with pictures of human genitals. Those community members didn't like the idea of a profile page to start with, and they were none too pleased with the result. They went to the Executive Director's user_talk page to explain why they were unhappy. They they thought it quite reasonable to illustrate their complaint with screenshots of their profiles. One screenshot had a pair of penises. The other screenshot had a lovely bleached asshole and neighboring vagina. Posted on the Director's personal user_talk. Lila handled it gracefully, although it would have been hard for her to blame the editors. That was the last I heard of the Profile project.

Well, that was the last I heard of the project for four years anyway.

Hi @Alsee -- thank you for checking out the notes from our offsite. I think you can tell from the page, but I do want to underscore that we're having just the earliest exploratory conversations around this idea. There's two main reasons that we've started talking about this:

(1) We've been talking lately about how a new user might be more likely to continue editing if they have an "identity" that they're developing. They're building their skills, building up their contributions, and making progress -- perhaps they want to show that to others. And in fact, we all know that thousands of users have elaborate user pages that explain about themselves and their work, and that that's important to them. We're going to try to learn more about this "identity" concept and how it relates to newcomers.

(2) When we ran user tests for the newcomer homepage, we saw that many newcomers independently proposed that such a space should exist on Wikipedia, in which they can show other editors who they are and what they are interested in.

So that's why we started thinking about this. Some of your links are helpful, especially to see community feedback on where the previous attempt at something similar went wrong -- that's a really good point about page images. If we continue to think about user profiles in the future, there are a few things I know for certain. We'll definitely be posting in public and encouraging discussion at every step of the way as we proceed (just as these notes were posted). We'll definitely be mindful of giving experienced editors flexibility -- we'll want any feature we make to allow customization at the wikitext level if possible. We'll definitely want to make sure that users have as much privacy as they need.

I've seen the claim that Wikipedia is not a social network from time to time, but the claim is wrong. Wikipedia is an extremely social network where the core purpose is to enlighten other people by providing crowdsourced content. A service that is non-social would provide no feedback mechanism at all, that is no discussion page, and would provide no way to cooperate with other users.

It would be more correct to say that Wikipedia (and the other projects) lacks some of the features being known from purely social networks, like a dating service, but from that to saying Wikipedia is a non-social network is a pretty far leap. I'm not even sure Wikipedia could exist without a substantial social part.

It is important for the users to be able to create and maintain an online identity, and such an identity is equally important on Wikipedia. Without a clear and distinct identity people will leave. There are two core factors on projects like Wikipedia when it comes to investments in identities, and that is cultural capital and social capital. Both are credited the user and the proxy (s)he has at the site – the user page. It must be more descriptive for the users, as the present user pages lacks personality. At best they are a complete mess.

@Theklan, @JVargas (WMF), @Amire80 -- we noticed (via @Tgr) that "parameterized user pages" is one of the draft recommendations from the Diversity working group. I wanted to point out this page to you, in which the WMF Growth team is in the early stages of thinking about a very similar-sounding project. We're considering working on it later this year, and hope to have any thoughts or opinions we can get. Hopefully it is a good sign that both the Diversity working group and the Growth team have interest in this!