What is a meta for?

Question

A guest at my house over the weekend stood puzzled next to my driveway. "What is that for?" she asked, pointing to a framework of 2x4s and hardware cloth perched on an old tire amid piles of sand. "Sifter." I said. "For what?" "For sifting gravel." "Why?" "To line walking paths, such as the one you're standing on..."

Spend enough time working with something, and it's easy to forget how strange and unnatural your makeshift tooling can appear to those unfamiliar with it.

Such is the nature of these meta sites. For 5 years, we've been directing folks to them for discussion and feedback, to commune with their community, hash out misunderstandings, propose changes and document the current state of things. And... We haven't always been good about helping them get past the initial shock and confusion of dealing with this system. This has manifested in several unfortunate behaviors:

• Downvoting of legitimate support questions - those that are not duplicates and reflect an actual bug or honest misunderstanding of how the system operates. Folks who are acting in good faith should be treated as such - we should be doing more to help them thrive, not shutting them down.

• Closing questions that seek input and discussion from the community - in hindsight, adding "primarily opinion based" as a close reason on meta sites was... a terrible idea. The old "not constructive" close reason made some sense for rants and so on, but for a discussion site closing down questions that solicit opinions is just silly. Here on MSE, we prefer the following off-topic reason instead:

  This question does not appear to seek input and discussion from the community. If you have encountered a problem on one of our sites, please describe it in detail. See also: What is "meta"? How does it work?

• The perception that Stack Exchange is disrupting the normal operation of the site by talking about future plans - even when those plans involve the community. I talked about this a bit more on Should Stack Exchange Inc promote its pet projects on meta sites?, but seeing this kind of backlash toward folks at the company who were trying to be open and receptive to discussion about what they were doing was kinda disheartening - we need to be doing more of that, not less.

• A reluctance to ask questions about asking questions - which is about as meta as you can get, and yet we probably get more of these via email than we see on meta, while being the least-qualified to answer them well.

After an increasing number of complaints from both core users and moderators on a number of sites, we're taking a hard look at how we can improve this situation... Starting with a change to the "What is Meta?" help center topic:

Meta is for...

• ...$SiteName users to communicate with each other about $SiteName (asking questions about how the websites work, or about policies and community decisions)
• ...$SiteName users to communicate with Stack Exchange the company (posting bugs, suggesting improvements, or proposing new features), and
• ...Stack Exchange to communicate with the community (soliciting feedback on new ideas or features, or discussing policies that affect the whole network)

Lots of topics are fair game, as long as they are about the Stack Exchange family of Q&A sites in some way. Any issue that concerns the core Stack Exchange engine that runs all Stack Exchange websites is welcome here.

This replaces the previous bullet list of specific topics, and aims to outline broad purposes with examples for clarification rather than constraining the scope to those examples.

So... that's a start. But a documentation change isn't going to change misconceptions that've already built up among newer users. So I'm asking y'all for help:

What do you think meta should be used for, and how can we better communicate this intent to new users?

Answer 1

A radical idea would be to change the terminology on meta. Not every question on meta is actually a question and answers are not always truly answers. So e.g. we would have inquiries and responses instead of questions and answers on meta (these are placeholder names; I don't actually have a good idea for real ones).

Meta is different, and making some obvious changes like this could help to avoid users applying the main sites rules to meta. It would immediately show that meta is different, and could help to clarify what it is for.

This is just an idea. I'm not convinced that it is actually a good one yet and haven't thought it through, but it is certainly an aspect that currently confuses users on occasion.

A fundamental issue is that meta has to serve two completely different groups: the users that know how meta and the whole network works, and anyone that uses the site and has some problem with it. A problem on highly active metas is that the community tends to apply the standards of the former group to everyone that posts, and that leads to a considerable amount of friction because that is simply an unrealistic expectation directed towards users that barely know how to use the site.

I might be okay with disabling downvotes for support questions. I personally think certain ones deserve downvotes, but my impression is that there are a bit too many undeserved downvotes on support questions. This would leave closing as the main action to deal with bad support questions, which might be enough.

Answer 2

TL;DR

This got a little longer than I expected, but the core of this suggestion is to attack the "meta" problem by trying to simplify meta for new users by highlighting the important content in the "FAQ" category and directing them to existing content. The goal will be to hopefully head off new questions that tend to get downvoted for lack of research.

By simply preventing as many of these questions as possible, you will by default make Meta a friendlier place.

---

One of the biggest problems I often see is the "lack of research" downvoting. I do it myself, so I am certainly guilty of this too, but I think many well established users forget what it is like to have a question in a new community and have no idea how to even find the answer.

A step to possibly help this would be to make the FAQ more accessible. The faq posts can be hard to find and are not consistent through all meta sites. The step you took on MSO was a step in the right direction, but I still don't think it does far enough:

There was a FAQ sidebar on the old Meta SO, in recognition of both the Community FAQ and how frequently some questions were re-asked. I've re-enabled that here, in hope of catching a few more people before they ask common duplicates.

Most establish users know how to find the faq posts, and know how to use it, but it is not intuitive for new users. Likewise, because the faq posts contain such a wealth of information, they can be difficult to filter through. In most cases, the faq posts are far more valuable than the Help Center itself because it does into far more detail and even explains why in a lot of cases. And then add in the fact that the answers to many support questions can be found in the faq.

All of this is compounded by the fact that the faq posts simply don't exist on a child meta until someone creates it. I realize that is in part by design because you want to encourage the new community to come up with its own guidelines and establish its own identity, but there are a subset of the faq posts that describe how and why the system works as it does, and these will not vary from site-to-site.

So I think the approach to making meta sites more welcoming is to help the new user more and do it in such a way as to make it impossible to miss or ignore.

I think there are multiple components to this solution and all together will make the meta sites a bit easier on brand new users, which should hopefully attack the problem from the new user end instead of from the established user end.

---

Step 1 - Make the FAQ more complete on all child-meta sites, but still give each community its own voice

Rather than a single faq tag, we should split it into a community-faq and stackexchange-faq (the names may need some work)

• stackexchange-faq questions would be the core of how and why the site works from a software level. This would/should not vary from site-to-site, with maybe a few exceptions due to special circumstances. Examples questions from MSE:
  • Why do I need 50 reputation to comment? What can I do instead?
  • Why does the Community ♦ user approve and reject edits?
  • The Complete Rate-Limiting Guide
• community-faq questions would be the communities own self-identity. Why do things work on this specific site the way the work? What kind of questions are acceptable here? These would be useful for understanding the community, Examples from MSO:
  • When to flag an answer as "not an answer"?
  • What is the proper way to approach Stack Overflow as someone totally new to programming?
  • Where can I ask questions that don't fit or are off-topic for Stack Overflow?

The "Stack Exchange FAQ" questions should be housed on MSE and either linked from the child meta site or be pushed to the child meta site when it is created. Edits should be allowed but edits to the MSE post should take precedence and be psuhed to the child meta posts in order to quickly and easily edit the FAQ to account for changes in the system. These need to get pushed to the child metas because established users need a duplicate target when these types of questions come up, rather than sending someone to MSE just to read a description of something.

The "Community FAQ" questions should be self developed by the community after the private beta ends and continue as the community begins to establish itself. Common questions that are relevant to that community (such as how to flag, how to review, what kind of edits are acceptable, what kind of questions are accepted)

Even if you don't want to split the faq tag, the most important part here is to seed all meta sites with a portion of the MSE faq questions to make the "user manual" easily available to all users, even those who only belong to 1 site.

---

Step 2 - Make the FAQ more obvious

The FAQ posts need to get top billing on every meta (both MSE and the child metas). I know they live in the "frequent" tab, but that just don't jump out at an unfamiliar user as being the "user manual" for a Stack Exchange site. It needs to be more obvious that it is the user manual. Users coming across from forums are used to seeing the "How to Ask" and similar questions pinned to the top of the question list. While we don't want to be a forum, we still need to make it easier to find for users expecting to find it there. Some thoughts on this are:

• Change the name of the "frequent" tab to something more obvious to the Text-Speak generation, such as FAQ or even just call it "Site Guide" or "User Manual"
• Giving them a button in the top next to the "Questions" button.
• The faq (or the 2 faq tags if Step 1 is completed) should be pinned to the top of the tags page
• Add a quick search of the FAQ questions. Could be something as simple as pre-populating the FAQ tag(s) in the search box if someone wants to search the user manual.
• Ensure that the FAQ "table of contents" is the top voted FAQ post or make sure it stays pinned to the top of the tag page (and even the tag search).
• Ensure all postings within the FAQ have a link back to the "table of contents" question.

---

Step 3 - Highlight the frequent feature requests page and direct users to it

New users always think they have the "greatest" ideas that no one has ever thought of before. They see a glaring need write a question and wait for a respond. Some even spend a lot of time carefully crafting a good feature request, complete with supporting info and samples, and then they run into a brick wall of downvotes. I think most MSE/MSO users know the biggest one: Should a user be forced to add a comment when they vote to close?

There is a lot of history behind this specific request (and some of the other most common "great ideas"). The established user community is tried of seeing the same feature requests over and over and over again, and the downvotes tend to rain down harder with every time it is asked. While some of the downvotes can be justified as the question itself may be poorly written, there should be a way to head them off by intercepting the question before it is even posted by giving the asker some idea of what the most common feature requests are and linking them to the previous discussion(s).

• Borrowing from On meta sites, require at least one non-required tag, suggest users add a non-required tag to help categorize their question when they use the feature-request tag. Because the non-required tag may not exist, it shouldn't be mandatory, but a strong encouragement should help the users who want to bother to read.
• Make a link to the "frequent feature requests" prominent when a user selects the feature-request tag on a question.
• Based on the selected tags, show the user the most common feature requests in that tag and strongly encourage the user to read them before they post their question.

---

Step 4 - Hit new users over the head with the FAQ to the point of annoyance

We all know that most new users don't read everything we give them, and often that is used as a reason not to implement an idea, but for meta sites maybe it is necessary to do something.

• When a user first accesses the meta site, make sure the FAQ is prominently displayed
• Make the "frequent" (or whatever it is renamed to) tab the default tab for users with limited meta activity.
• Prioritize FAQ questions in the "Your question may have an answer" list when asking the question, especially for user below a certain rep.

Answer 3

Why does the about/tour page not mention Meta?

Either include it there, or make a parallel page (with an EXTRA bronze badge! yay!) on the meta site itself. Or both.

Answer 4

It would be a good idea to encourage new users, who might not be aware of Meta sites, to visit them with questions about the site, especially if they don't understand why their question has been closed/downvoted (invariably, this gets cleared up very quickly in Meta, where in comments it could be bogged down by discussion and other valid comments).

Certain close reasons, such as off-topic or site-specific reasons, might benefit from links to Meta, so that newer users don't get confused.

This, combined with a good FAQ specific to Meta boards, would improve the quality and output not only of the Meta boards, but of the main SE sites as well.

Answer 5

I think the three bullets in the new Help Center description pretty much covers the "What do you think meta should be used for" part, so nothing to add.

As for "how can we better communicate this intent to new users", I suggest the long time "hard core" members of each meta site (MSE included) will do a joint effort and restrain themselves from casting downvotes too quickly and/or posting snarky comments. I admit falling to this myself sometimes, but I am also willing to try and improve. When new users who decide to stick in the site and start being active would see the veteran users acting polite and not downvoting valid questions out of spite, they will follow.

Can't think of any substantial change in the system itself, but something we can do and currently is not done enough is let users who act wrongly know they did something wrong, maybe linking to this very question.

Answer 6

Mad Scientist makes a good point in his/her answer. If Meta is different, there should be different features to incentivize users acting independently from the main site.

I would focus more on improving communication about 'What Meta is for' to all kinds of users rather than the new ones. New users should inherit good behavior from "elderly".

Here is one suggestion: create some badges specific for Meta [e.g. has positively scored answers in 10 consecutive days adding a minimum +2 score for at least three of the main four meta tags (discussion, support, bug and feature-request), under questions which are not closed and that were not (re)tagged by yourself ---> community-manager (silver)]. The suggestion here would be to open an exclusive post asking for opinions/ideas.

Answer 7

tldr; this is essentially in support of status quo

From my understanding, Meta is for discussing issues that arise during the use of the system which should not be placed at the point of encountering them.

How as this recently changed? Not sure it has.

These large meta discussions seem to arise from mSO posts. The major change this year was to make mSO a clean slate. I believe that every once in a while discussions which have been re-hashed so many times before are posted to mSO. However, they are no longer at mSO, they are hidden in the history of MSE now.

The result is a clash between those who remember the original hashing of the issue, along with all the intense back and forth they tend to generate, and those who find legitimate points in the current situation highlighting said issue.

Instead of re-hashing the issue at hand, legacy users down vote, because they know this has already been discussed. Unfortunately, since it is hard to find where, there is not really a good way to send newer users to the original discussion so the downvote is all they get. They feel wronged because they got no response to a legitimate issue.

This circle of old and new discussion is destined to continue. The obvious reason is that as posts age their relevance wanes. But the current reason is that mSO still lacks tens of thousands of posts to support a community of millions. These jarring reflections will continue until the balance between mSO content is more in line with the level of content on Stack Overflow. I believe that number is probably in the 75,000 to 100,000 posts range. Right now, there are only 6800.

So expect to see more of this! When there was only a handful of posts on mSO people were discouraged, mostly because they asked age old questions but did not know it. Now users are finding all of the hidden gems and asking them.

There is no preventing this, and the only advice I would offer is to be ready to answer these questions when they are asked on any metas whose content is greatly lacking. On mSO, we still have a long way to go.