The 2022 Community-a-thon has begun!

Question

Our third Community-a-thon kicked off today. This is a three week event that will continue through November 8th.

This is an event we hold for the staff at Stack Overflow. During the course of the event employees are encouraged to participate on sites in the Stack Exchange network, in order to increase their familiarity with the public sites and experience things from the perspective of a community member. This is the third year we've held the event and you can see our previous posts from 2020 and 2021 for details on the past Community-a-thons.

The goals for the event are similar to those from the previous two years:

1. Improve Empathy between all Stack Overflow employees and our sites and communities.
2. Achieve a high degree of participation during the event, that can hopefully carry on afterwards as well.
3. Increase our collective familiarity with our core product and elicit useful feedback on actionable ways to improve our functionality and user experience.

Similar to previous Community-a-thons, the event will include a competition in which participants can receive points based on engagement and activities done on network sites during the contest. The point system will be very close to what we did last year. Like last year we will not have a dedicated chat room for the event but we are letting Stackers (what we call Stack Overflow employees) who are less familiar with the Stack Exchange network know about chat so don't be surprised if you see some newer Stackers in The Tavern or in some of the site-specific chat rooms.

Also similar to the last two years, some Stackers may be participating with their staff accounts while others may be using brand new accounts. We're encouraging staff to use new accounts on sites where they have a lot of rep so they can remember the experience of what it's like participating when you haven't yet earned a ton of reputation. Those participating, regardless of whether they are using newer or established accounts, are encouraged to submit their feedback, observations, and experiences over the course of the event.

We are sharing this here to keep the community informed that this is a tradition and effort we're keeping alive. Community is the core of Stack Exchange and we want to make sure that all employees are continuing to improve their familiarity and empathy with the network.

You gave us a lot of great feedback last year about your observations around staff participation during the 2021 event and we used that to help craft and info session we're holding for Stackers who are less familiar/active with the Stack Exchange network. Members of the Community Team, Product, and other Stackers who are on the sites frequently are around during the event to help newer members out if they have questions.

If you have any questions related to the event or its goals feel free to ask below.

Answer 1

Last year I was expecting that someone from staff write about their personal experience in the Community-a-thon. Nothing extensive, just a couple of paragraphs about what they had learned and how they enjoyed the experience.

I know this might involve a staff member confessing that they acquired new skills, something like:"I didn't know how to edit very well, but I did 50-100 edits and learned a lot from the experience." Or: "I tried close votes and reviewing, and the experience was initially challenging." Or even: "I participated on a smaller site about a subject I hadn't had the time to learn more about."

So I was a little disappointed last year that no one got to share their experience report.

Answer 2

We're encouraging staff to use new accounts on sites where they have a lot of rep so they can remember the experience of what it's like participating when you haven't yet earned a ton of reputation

For a true user experience, I'd recommend only using regular accounts and not accounts with special flairs (staff/mod/etc.). E.g., close voters and downvoters are much more timid if the OP is marked as staff/mod.

This is a three week event that will continue through November 8th.

Note that 3 weeks won't suffice to experience some features, e.g. Roomba.

Answer 3

Sharing my observations as a high-rep user pretending to be a low-rep user.

I've seen this network from all angles: a new user almost 14 years ago, a high-rep user, a moderator for several years, and now a staff member. This exercise gave me a new perspective as a low-reputation user on Stack Overflow which is, arguably, a much different place than it was in 2008. I shared some observations internally but I'm happy to share a few of those - slightly more sanitized - here as well.

Earning reputation is harder than it used to be.

Certainly it feels like it is harder to earn reputation as a new user these days, across the three main ways to earn it: questions, answers, and editing.

Questions

In fact, on the tags I frequent, questions on the front page typically average a negative score at any given time. The informal correlation is that newer users are asking most of the questions, and maybe they're not asking great questions, but something not clear from the data and the votes alone is why? A thought I've had over the past several years is that there are fewer good questions to ask that haven't already been asked, and it's easier to down-vote than to find a valid duplicate. On the other hand, most of the questions that I actually down-vote are from users who aren't brand new, but also clearly haven't taken the tour, or followed previous prompts to provide more/clearer info, or bothered to understand how to make even a question they didn't bother searching for fit the bounds of the site.

Answers

I did some modest analysis on my own answers from a low-rep account because it somehow felt to me that the answers were received less well simply because the person posting them didn't have high rep. And there were certainly a few cases where it seemed like the votes went to the higher-rep users regardless of the actual answer content. Very subjective, sure, and I'm not going to shame anyone, but there was more than one time during this experiment where a higher-rep user posted a wrong answer - in one case it wouldn't even compile - or ignored crucial details in the question, but they still received multiple up-votes even when other users pointed out the problem(s).

Since my main account has multiple answers with three-digit votes or more, granted over a much longer timeframe, it felt like it was very hard to come by even the sniff of a single up-vote this time around. Looking closer, you can see that I only had a single answer with more than three up-votes, and 28 answers with a zero score. I didn't correlate everything together in a matrix, but it was certainly interesting to me how many of my answers were accepted with 0 votes, and how many OPs accepted my answer but didn't up-vote it themselves. I should have included how many of these did not accept my answer or any other answer, but I'm not going back through them a second time.

With such a modest sample it's hard to draw anything super conclusive from that, but it reinforces the feeling I have that people are stingier with their up-votes, reserve their quota for higher-rep folks (or simply names they recognize), or perhaps my low-rep account rubbed all of these users the wrong way.

Editing

Many of my colleagues in this experiment noted that - especially on smaller sites - it was very hard to earn reputation from questions and answers because coming up with a new question that is received well (or coming up with a new answer on an old question) is very challenging. This goes along with my feeling that there aren't a lot of great questions that haven't already been asked (again, in the tags I'm in, anyway). So one thing folks do to try to gain reputation is suggesting edits. As a new user, this is unbelievably frustrating, because until you reach 2K rep and can edit freely, most of the time you're met with this:

Even worse is when the queue has room for you when you start your edit, but has filled up by the time you finish, or your edit has simply been trampled by another user.

My suggestion here was to have a lower rep threshold for editing, like maybe 200 instead of 2K, or have it simply not tied to rep at all. (Like with questions and answers, we've all seen high-rep users make bad edits, so it seems we artificially assume that earning rep makes you a good editor. And as a low-rep user but with high expertise in a topic area, being able to recognize but not correct a typo or incorrect tag or until 2K seems so. far. away.)

Admittedly I haven't fully thought this through, it just struck me as something that could be improved with the right heads involved. Maybe you get a three strikes policy on suggested edits; if you can get to 5 or 10 or n accepted without three (or some n) rejected, you earn the right to edit. If you earn the right but then have n edits rolled back, you start over and need to earn the privilege back again.

And to allay fears of an overloaded queue, maybe have bigger rewards, encouraging more participation, for reviewers on the suggested edit queue who perform some # of edits without failing audits. Much of this workflow is already there, so we wouldn't be inventing anything.

Other random feedback

The volume of off-topic questions here

I pointed out that the Ask Question button here on meta starts off - but only to new users - with a rather harsh demand, telling them to go elsewhere:

It isn't clear from the wording "Ask your question on Stack Overflow" that they have any choice. Based on the volume of nonsense that gets posted erroneously to meta anyway, it is clear that most people completely ignore the popup (or have managed to surpass the rep threshold without actually learning what the site is for). I think the wording in the dialog could be improved; it just came across to me as jarring that we're immediately sending people to go ask on Stack Overflow, and assumed that if they're asking a question in the wrong place that the right place could only be Stack Overflow. All of this before we've given them any context about why they might actually be in the wrong place to begin with (which, admittedly, does happen later in the dialog). Reorganizing the content and perhaps forcing users to scroll to accept (a la some other software) might reduce that volume by a bit, but for most it is probably fruitless to force it that way.

Another idea is if a user of any rep gets a question closed as the common and specific off-topic reason, they should continue seeing the dialog for their next 5 questions - even if they have clicked "don't show again."

Ability to view source/markdown

A limitation of not having carte blanche to edit is the inability to view a read-only version of the markdown of a post. This is useful not just for learning how markdown works, but also for copying tabular data, code, or anything where it is useful to retain the original formatting.

About half of the times I click Edit, it's not to actually make a change to the post, but rather to get at the markdown to construct a repro. This usually involves sample data in a markdown table that is much more directly usable than plain text (or view source -> HTML).

The current behavior - where the edit queue is often full, blocking the textarea from showing - is a significant workflow blocker. Multiple times, I had to switch to my high-rep account to copy and paste a post's markdown. Most users don't have a magical high-rep account to switch to.

Yes, there are round-about ways to get there (example), but there should be something fewer clicks away (e.g. one).

So, the request here is pretty simple: another button or link somewhere around a post that says "view markdown" or similar - so that all users have access to the markdown, and so that Edit stops pulling double duty.

Popup warnings on tag combinations

New users frequently apply multiple tags that are usually mutually exclusive. The tags are conflated by the user because they don't read the tag descriptions and/or are trying to tag spam in order to cast a ridiculously wide net.

The combo I see dozens of times every week (and which are now much more obvious with my recent burst of activity) are mysql and sql-server. It's pretty balanced in either direction, so I don't know if some people think their SQL Server instance is "my" SQL Server or if others think MySQL is also "a platform that serves SQL" - but these questions always end up having experienced users scold the asker and tell them to correct their tags. Quite easily 99% of the time, the user tagged one or the other "by mistake" and either fixes it themselves, or someone else cleans it up based on their response. Or they don't get cleaned up at all (sample any of these questions and you're much more likely to find that the OP didn't respond to the prodding, as opposed to the question turned out to really apply to both platforms).

The feature I envision defines sets of usually mutually exclusive tags and, when a user tags more than one tag in any given set, the UI prompts them with "Are you sure you want to tag both/multiple?" with links/hovers to their respective tag wikis.

This shouldn't be a full-on block because there are rare exceptions where tagging both is appropriate (e.g. help me with this specific issue I'm having migrating from one platform to the other). So it would behave like "Are you sure you want to post a second answer?" - which allows you to proceed, but tries to make you think about it first.

I have some more, but I think that's a good start for now.

For @ColleenV, there is expansion that happens automatically in posts (but not in comments I think):

• The 2022 Community-a-thon has begun!

The markdown that produces the above line is literally:

- https://meta.stackexchange.com/q/383022/

Answer 4

Can we please have statistics in the end of the events?

Partial list of what interests me:

• Total number of staff members taking part in the event
• Total number of Q&A sites taking part, i.e. on how many sites they participated
• Total number of posts made by staff members as part of the event
• Votes summary (i.e. upvotes & downvotes) on the posts made by staff during this event
• ...and more of the same?

While it won't really measure success or failure of the event, I find it interesting things to know.

Answer 5

Here's a bit of an overview of my experience participating in Comm-a-thon:

Overview

I've been a user of Stack Overflow for almost 12 years, and it's been a long while since I didn't have enough rep to pretty much fully participate on the site. And for the last few years, I've been a dev at Stack, so have full privileges on the site, and see things rendered on the page that others don't (like page performance metrics).

For this Comm-a-thon, I used a low-rep account I created a couple years back so that I could see the site as a non-privileged user, mainly to reproduce/view reported bugs that aren't visible to a mod/staff account. At the start of Comm-a-thon, this user had 1 rep on SO and another SE site.

My approach to Comm-a-thon was shaped by these factors:

• As a developer on the team that focuses on Collectives on Stack Overflow, I wanted to experience the Collectives as a new, non-privileged user. I spent a fair amount of time just going through the mostly-readonly pages related to the CoSO product.
• As someone who mentors young (school aged) people interested in learning coding, I wanted to use the site as if I was interested in participating in SO, but didn't necessarily have the knowledge to answer complicated questions. This meant attempting to gain rep and participation privileges without answering questions that an inexperienced dev wouldn't know.
• I primarily use Chrome as my browser and I wanted to use a lesser-used browser to experience the site, so I utilized the Edge browser.

I set aside time on five days to specifically participate in the Comm-a-thon, spending between 20 min and a hour each time. Most of my effort was focused on stackoverflow.com, but I also utilized meta, Gardening & Landscaping and Software Engineering, including asking questions on those other sites.

Hard to earn rep

Much like Aaron described, I found earning rep to be very difficult, especially since I was intentionally avoiding answering non-trivial questions. Because I had generally no privileges on the site, participation was extremely limited (no comments, no voting, effectively no editing).

The biggest frustration point for me was the inability to edit due to the edit queue being constantly full. On all five days, I tried to edit posts to improve formatting, or edit tag wikis that were missing. I ended up submitting a total of nine suggested edits: six improvements to questions and three tag wiki suggestions. All but two of my suggested edits were accepted, with one tag wiki being rejected because I included markdown in the wiki excerpt, and one tag wiki edit that is still in the review queue seven days later. I was surprised to discover that when a review is rejected, even though the reviewers left comments as to why the review was rejected, I got no notification from the system about the rejection or the reasons. I only know about it because I went looking in order to write this post.

The question edits I was able to submit were reviewed fairly quickly, with four reviewed within minutes, one in about an hour, and one after eight hours. My tag wiki edits, however, took a long time, with one getting rejected after five days, one getting approved after seven days and one still in review queue after seven days.

The worst experience was when the system would allow me to type up my edits, but then refuse to allow me to submit them due to the full edit queue.

Additionally, I reported a UX bug where the edit wiki page for a tag without an existing wiki says "All registered users may propose new tag wikis. (Note that if you have less than 20000 reputation, your tag wiki will be peer reviewed before it is published.)". BUT... there's no edit button or any way for me to submit edits. The edit button is hidden when the review queue is full, but there's nothing on the UI to indicate that.

Ultimately, I ended up reaching 15 rep on SO just today, which finally gave me rights to vote and flag posts, but I still can't comment, which greatly impairs my ability to participate broadly. On the Gardening & Landscaping site, I ended with 11 rep, and the other sites I wasn't able to gain any rep while playing the role of a junior dev.

Ask Wizard leads to some odd questions

As I was looking for questions to edit, I came across a pattern that I attributed to new users using the Ask Wizard. In the Ask Wizard, the UI presents two questions: What are the details of your problem? and What did you try and what were you expecting? Both questions require a minimum number of characters to be entered before you're allowed to continue. These answers from these two questions are then combined to become the actual question posted on the site, but we don't include the prompt questions in the output. What I found was several questions which had an ok entry paragraph, but then had a final sentence like this:

(nothing really "tried" or "failed" - purely a documentation question)

Nothing Nothing Nothing

So the user had entered an answer to "What did you try?" that, out of that context, provides no value to the final question. They entered text purely to get past the minimum character requirement. I provided this feedback to the appropriate internal team for consideration around other potential ways to prompt the user, or maybe not force the asker to answer the second question.

UI/UX unexpected behaviours and bugs

In the course of using the sites, I came across a few odd UI behaviors that I reported to the appropriate teams. One dealing with spacing around an Ad when a question is migrated to another site, some dealing with UX on our internal Stack Overflow for Teams instance (which we used for collecting feedback during the Comm-a-thon) such as the unexpected inability to post animated gifs, and some pages hide the left-hand navbar even though my preference is to always show it. Most of these are "known issues" that just haven't been dealt with yet.

Additionally, in the user profile bio for my low-rep user, I have a link to my "real" user. I was surprised to find that when I created an account on a new site, while my bio was copied over, the link was stripped. This is a SPAM account deterrent for low rep accounts, but was unexpected.

Answer 6

I'm glad to hear that y'all are doing this again! I've always disliked how the staff of other forums (looking at you, Reddit) have little to no familiarity with their subjects forum users.

On behalf of Worldbuilding.SE, I welcome our corporate overlords.

Answer 7

Would it be possible to post an anonymized summary of the staff activity and comments publicly maybe annotated with lessons learnt and possible future actions, if any?

Also a suggestion for a possible future exercise would be to take a group of people with genuinely no experience of the SO sites and community and perform a similar monitored exercise - possibly organised and recruited via appropriate college or adult education course tutors, ideally in several countries. This would give SO some real new user feedback while benefiting the students (I am assuming that most tutors would feel that the relevant sites would be a positive element in the students experience). I am not sure what incentives SO could offer to encourage participation, but I don't expect that it needs to have too great a financial element.

Answer 8

I generally don't like the idea of sock-puppet accounts, but I can understand why they'd be necessary for this useful exercise.

---

What happens to these brand new accounts at the end of the Community-a-thon:

• are they merged with the staff members' regular accounts?

• or are they just abandoned?

Answer 9

After some thought, I would add that (for the purposes of the Community-a-thon) if a staff member is about to do "mod-like stuff" on a particular site (i.e. edit without review, closing, etc.), that maybe they (first) contact the existing mods in the mod chat room (of that particular site, or, if need be, in the TL) to notify them of their intentions..?

A more personal form of headsup and/or self-introduction might be a good idea.

This is related to my post on the Stack Moderators site.