What are the effective communication channels for effecting change to SE?

Question

It seems clear at this point that MSE is not a useful mechanism for pointing out problems on SE sites and getting resolution for them.

It is useful in some ways. If SE staff ask questions about future actions being undertaken, there is a non-zero chance that they will pay attention to some of those responses and take them into account. While the new site layout still did happen despite lots of people not liking it very much, the powers that be were responsive to specific issues about certain aspects of it. So clearly, if SE staff ask for feedback, there is a non-zero chance that the feedback may matter.

Also, posting bug reports on MSE does sometimes lead to those bugs getting solved, in many cases.

But when it comes to non-bug issues with Stack Exchange as a network, it has become abundantly clear that asking a question on MSE will accomplish nothing. There are many lines of evidence for this.

One of them being that SE staff have explicitly stated that many of them don't care about what gets said on Meta at all:

Often (and more often recently) I've heard colleagues dismiss meta feedback.

Another line of evidence is a more recent event. There have been dozens of MSE posts about the "Hot Network Questions" list, many of them speaking quite negatively on its effects. These range from issues about how it causes a flood of users due to detecting controversial questions rather than good ones, to having to see a list of question titles that seem... less than professional. Things have occasionally been done, but these have been pretty minor. Regex filters for certain keywords and the like.

However, a few Twitter posts managed to find their ways into the eyes of a SE developer. This user managed to get approval to exclude IPS from the HNQ list, then implement that exclusion, and push it out into the live network.

Within one hour!

Equally rapidly, a post appeared on MSE, detailing a clear intent from the SE staff to fix the HNQ list or, if it proved unfixable, to scrap it entirely.

Now that is what I call an effective communication channel. In a scant few hours, SE company policy on the HNQ shifted visibly and dramatically, from an apparent course of "meh, whatever" to "OMG, we gotta fix/ditch this!"

My overall question is this: what communication channels can we use to get similar response times to user problems? Obviously, all the Meta posts about HNQ's problems had little effect, compared to a few tweets.

So, do SE users need to become more active on Twitter? Who should we be tweeting our issues to in order to get them heard? How should such tweets be composed?

Do SE users need to be making blog posts, and if so, where? What would be the most effective means of catching the attention of SE staff, such that our concerns will be taken seriously?

Note: This question is not about lambasting SE for not listening on Meta. That's a ship that has already sailed. So feel free to take comments on this subject elsewhere. I'm genuinely asking about what the best way to successfully communicate important issues to SE, such that they will actually respond to them.

Answer 1

Meta isn't scaling well for any of the goals that it was originally designed to meet. I don't want to get lost in discussing all of the ways that it's just not working; I'd rather just say that our structure when it comes to development evolved considerably and meta simply didn't.

I don't have a single good answer here. We're way past the point where the person out in front engaging with the community has direct control over priorities and what gets worked on next. We just can't delight people like we did in the old days where we'd roll out new stuff sometimes the same day it was asked for. I have substantial influence in this area, but ultimately, the PM that owns the product has to be able to define and manage their priorities. And those priorities aren't always open to negotiation. I hate to say it, but revenue really matters. That's not to say what you want doesn't, but we're not always able to talk about stuff that takes precedence or why. We didn't have that problem when our only source of revenue was ads and everyone knew who our customers were (well, everyone that disabled adblock).

Things happened the way they did this time because we positively had to do something about those titles being shown out of context, and there was something we could immediately do. Had the complaint have touched on something that required substantial resources to fix, the best we probably could have offered is an acknowledgement of the problem and a rough guess on when we might be able to get to it.

Meta doesn't track bugs well, it doesn't track feature requests well, it doesn't indicate presence from us well (how do you know what stuff Joe Friend has been looking at recently for distant scheduling?) and quite frankly, even working 60+ hour weeks, there's still not enough time to get everything done so adding even more writing on top of the PMs list of things to do isn't going to help.

There's also no good way of keeping a roadmap up to date, or even tracking progress toward goals in any kind of public way.

I think, in 2019, we're going to have to bite the bullet and replace at least bug tracking and feature requests with something else, or have something else behind the pipeline consuming meta and making sure it gets put somewhere else in ordered form so we can work from it. I don't know what that's going to look like, I just know what we've got isn't working, and breaks even more embarrassingly by the week.

And I know a lot of you have got "well why don't you just ... " suggestions at hand, that's great, but they're probably not going to help because you really need to be behind the scenes to know what cogs aren't going to fit.

So, for now, use Twitter if Twitter is comfortable for you. Use email if you want to use email. Use meta (many people don't because it's scary) - it's not that we're deliberately ignoring things, it's that the pile is so huge and continues to get so every day we pretend Meta still works.

The medium you use to talk to us isn't going to significantly impact how we resource things; the biggest factor there is what we might be able to do with whatever free resources are at hand. We need to replace this ad-hoc opportunistic game of fix-a-thing with systems specifically designed for modern development (don't worry, we'll still keep it around for discussion), but I don't have an ETA for that.

In short - I can't guarantee anything is going to be more successful than what we currently have, it would depend on the urgency of what you were reporting and how quickly we could mobilize to do something about it. Getting more eyeballs or people yelling about it will certainly get our attention, but getting our attention is in no way a guarantee that you'll get what you're after.

We know this is broken, and it's something we need to fix. And part of that fix really is adjusting expectations - we just can't ship the stuff folks ask for nearly as fast as we could previously -- we're just too big and there are too many dependencies and stakeholders.

But, I'm going to put pressure on setting priorities to settle on something that actually works so that we can at least set expectations consistently and accurately.

That's as honest as I can be. We need to keep how open and responsive and transparent we are as we transition into being a more mature company with structure. Until we do that, it won't matter where you go.

And, well, that's what I'm doing.

Answer 2

Thought experiment

Let's imagine that this week, the same person with the same attitude had posted the same complaint on meta instead of as a tweet. What would have happened?

Feel free to post your own interpretation in the comments, but my answer to this is the following. Note that I'm not commenting on whether or not this is what should have happened, just that objectively I think it's what would have happened:

they'd have been downvoted to hell, probably closed as duplicate, possibly suspended for rudeness. Nothing would have come of it.

What can we learn from this?

• When people post civil and constructive suggestions on meta, as likely as not they'll be ignored. (Of course SE does implement some changes, and I'm sure they try to prioritise as well as they can given whatever constraints they have, but it's a fact that many nice and well-worded feature requests with positive community feedback have been status-ignored for years.)

• When people post angry and non-constructive rants on meta, they're almost always met with downvotes, close/delete votes, and potentially (depending on the amount of anger expressed) bans from the site.

• When people post on Twitter, there's no way for SE to moderate them or control the reaction. The only thing they can do to contain it is to actively and publicly respond. If they do nothing ... well, maybe the tweet will get ignored, and maybe it'll go viral and SE will lose all face and backers.

  Also (not re this specific tweet, but a general point) drama gets eyeballs. Controversy is much more likely to go viral than a reasonable, well-argued, balanced criticism that takes all facts into account. Or, as Eoin Colfer channelling Douglas Adams wittily put it:

  Guide Note: Throughout recorded history the ability to ‘state one’s case well’ has generally had about as much success as ‘talking things out reasonably’ or ‘putting aside our differences’. The people who use these tactics generally mean well and would make excellent motivational speakers or kindergarten teachers, but on no account should they be put in charge of situations where lives are at stake. [...] General Anyar Tsista, the acknowledged prince of negotiators, once claimed that while on the job he never used a sentence that did not include at least one zark, two shits and half a dozen asscracks. His final pronouncement contained only a single shit, and was uttered in the form of an authoritative command to his bowels, which had locked up as a result of too many hours seated around the negotiation tables. Unfortunately, because of their thin bowel walls, Golgafrinchans are prone to catastrophic bowel ruptures, so General Anyar Tsista’s final utterance was also what killed him.

From a cynical point of view, a Twitter rant might put SE in a position where they have to do something about a problem. A meta post never will, because they control the platform.

Answer 3

I don't see any reasonable way other than SE making an intentional effort to listen more to the community on the various meta sites, and follow through with it. Or just declare that meta doesn't matter, and only take bug reports and support requests on it. Good meta posts take a lot of work, if that doesn't lead to anything we should stop wasting it.

The correct place for feedback is here, connected to the site, from the people that are using it. Putting it on Twitter means the people with the most followers, or the best direct connections to SE count, and everyone else doesn't. The same counts for blogs, I don't have a blog, and if I had, I doubt anyone would read it. So only people with an audience matter.

The easiest way to systematically improve this would probably be to declare an arbitrary score threshold and task employees with looking at meta posts that meet this threshold and organize a response. This is a very old idea, and a rather blunt instrument, but organic responses by SE just don't seem to properly cover meta, so a formal way could help.

Answer 4

This is basically another form of saying

How and where can I generate bad publicity for {company X} such that the bad publicity drives them to change {thing that needs to change}?

To pick a big, dramatic example -- if the New York Times posts a huge sensational expose about a company, it's very likely to result in changes at the target company.

In both cases, I'm personally glad that the Twitter outrage happened because the underlying changes were ~5 years overdue and discussed extensively on meta beforehand to very little practical effect. Whatever the vector, at least now real action is being taken, and changes are being made, not just discussed ad nauseam. I guess you'd say the ends justified the means.

Is the lesson

Should I try to generate an intentionally sensational example and post it on social media in a sensational way, in order to trigger positive changes that have been discussed ad nauseam for years, but no action has ever been taken?

... well, kinda.. yes, you should? It works.

As long as your heart is in the right place, and your goal is truly and honestly to effect positive change (particularly the kind of positive change that has been discussed for years on meta in great detail with very little practical change), I for one don't have a problem with that, speaking for myself personally. YMMV.

Answer 5

Title of your question kind of hints at an answer. One of communication channels to consider for effecting change to SE is to discuss this change in the very same social media channel where it originated from.

Specifically, when Stack Exchange team refer or respond to particular social media post to announce or justify some change in the system our community members can get to that post and discuss it and involved change over there.

---

One caveat, authors of these originating posts might sometimes get unhappy about received feedback from our community. I heard that was the case with two recent changes ("welcoming" and blocking IPS from hot questions).

But I guess it should be responsibility of SE employees to consider consequences when they choose to hijack particular social media post and turn it into change announcement / justification.

Answer 6

I had been planning to be gone from the site until January, but having found out about recent events, I decided to drop back in briefly in the hopes that my input may help the site.

In theory, this should be the place.

But as a friend of mine used to say (God rest his soul).

"In theory, theory and practice are the same, in practice they are not."

Without rehashing what's been said too much.....

In order for any feedback venue to be effective, it has to do the following.

1. Address grievances

2. Listen to criticism

3. Acknowledge concerns

4. Demonstrate that it is doing 1, 2, and 3

As the recent dustup has demonstrated, if this is not the case people will exercise other options. I found out about it because it's made it to Facebook now.

It looks like people are upset and are going to make their voices heard.

It would be better for SE if those concerns were aired in here, and not elsewhere. The only reason people would go elsewhere would be if they felt like they were not getting satisfaction here.

So, the owners and mods at SE need to ask themselves the following questions HONESTLY

1. Are we doing everything we can to listen?

2. Are we showing the respect we want to receive ourselves?

3. Why would anyone want to go offsite to discuss what is going on in here?

I can briefly point out what I've seen that I think accounts for some of the frustration:

• If someone brings up a point and they are polite, they are often ignored.
• If someone brings up multiple points, and are polite, they are accused of sea-lioning
• IF someone gets frustrated, and brings up a point forcefully, they are called rude
• If someone gets REALLY frustrated and brings up multiple points forcefully, they are banned.

There's more, but I think that sums up my point.

I suspect that this will not be well received, but I do hope to be proven wrong.

Answer 7

It’s here it should be, but SE team need to improve.

As such in IMO;

I would think that after X votes, an official reply should be made by the SE team to show the request acceptance.

After a bounty in one of those request, an official reply with a status update should be made.

After X positive votes, after a status-declined was set on a request, it should be rethink and a status update should be written in it. As something not imaginable 5 year ago is maybe possible now.

In SE they started to give status back, they should talk about the top 5 requests still open there too, to give a status without having to offer a bounty.

Answer 8

Obviously, those channels that lack all the quality control and boringness MSE provides. Because it's not about lists of pros and cons, or about rational suggestions, but rather, as Jeff succinctly put it, it's about sensational examples which garner attention. And he's completely right - it works.

Use what has been a hot topic as an opportunity to learn: HNQ. More precisely, its flaws. HNQ draws attention (sometimes too much) with subpar selection algorithm which favours controversial questions instead of quality ones. Positive feedback loop around it makes those popular questions even more popular. Likewise, a well-followed social media account is a soapbox for anything you have to put out - the more flashy, controversial, rude, snarky, it is, more attention it gathers. The more attention it gathers, the more attention it gathers. Therefore it will get to someone at SE more quickly and will need to be put out accordingly quickly. I argue it doesn't matter whether that's Twitter, Medium, or an editorial in NYT. The only thing it matters is that it appeals to the masses. In other words, think like a populist, not like a meritocrat.

Don't get me wrong, hearing outside criticism can be tremendously helpful. It's a great resource to see what is being done wrong, a unique opportunity to escape the echo-box SE can become with established community. But instead of using meta sites as tools to prevent fires getting out in the first place, and when occasionally some slips out to contain the damage and put it out in a controlled environment, they are completely ignored by the higher ups.

There's no "prevention", so forget about it. Start fires, as they are the ones that need to be put out. It will work, because it has to work. Well, at least as long as there are less fires than firefighters (and considering response times, we're still pretty far away from that moment). Just be sure to get a big enough soapbox to throw matches.

Answer 9

Maybe the best solution is to close MSE forever?

[Emphasis in bold mine]

Starting from the OP's introductory paragraph

It seems clear at this point that MSE is not a useful mechanism for pointing out problems on SE sites and getting resolution for them.

It is useful in some ways. If SE staff ask questions about future actions being undertaken, there is a non-zero chance that they will pay attention to some of those responses and take them into account.

It also seems that the OP does not exclude the remotest chance, however rare that may be, that someone from the SE Team takes notice. They continue

While the new site layout still did happen despite lots of people not liking it very much, the powers that be were responsive to specific issues about certain aspects of it.

This bit is actually true. Some bugs and defects were fixed and some design features were improved. The OP opines...

So clearly, if SE staff ask for feedback, there is a non-zero chance that the feedback may matter.

Oh, wait – they meant that there is a chance of that happening now (see comments below)

Also, posting bug reports on MSE does sometimes lead to those bugs getting solved, in many cases.

Ah, good! I think…

But when it comes to non-bug issues with Stack Exchange as a network, it has become abundantly clear that asking a question on MSE will accomplish nothing. There are many lines of evidence for this.

But there had been requests and suggestions to improve certain features and aesthetic qualities of the new theme, the new theme which I, personally, dislike intensely. Developers, community managers, and product managers, responded and handled those requests when they could.
Don't believe me?

Examples of when SE staff listened and when ‘stuff’ changed via MSE

By Jon Ericson♦, community manager, posted on August 13

1. In between we see criticism with varying degrees of utility for us. For instance, we already know people don't like the new left bar. It's a central motivation for imposing the framework on sites and won't really prove it's worth until we are able to add features such as custom question list notifications to it.

By Joe Friend♦, product manager, posted on August 22

2. An occasional problem with comments and answers on meta is that they seem more interested in scoring points (actual and metaphorical). Instead of responding to the post, they speak to others who are equally upset about whatever product/UX change is being made. The result is that the commenter uses over the top language, denigrates the changes and or the people who made them, calls into question their abilities and shows general contempt for my team.

This appears to be an objective assessment by someone who was acutely aware of users' complaints. The SE employee is unhappy about the current hostile climate, but they seem willing and eager to work with the user base.

My team is committed to improving Q&A for you and all our users. If we can let some of the stuff from the past go and work together to develop mutual trust, then we can collaborate and make significant enhancements to the sites that you love.

By @Journeyman Geek, SE user and moderator on SuperUser, posted on August 13 [emphasis mine]

3. I personally love the current theme we have, and it's only because we engaged and spoke up and let folks know what bothered us. There's a few things that we did that worked great: • Start complaining early: go through the mockups with a fine toothed comb. SU lost a lot of the design elements we disliked because we noticed and spoke up against them […] We didn't like the 'filler' background on ours and explained why..

Overall, the user appeared to be happy. Were not his complaints heard and acted upon?

Back in March 2018, @Adám asked

4. Can we at least select our fonts?
   I think the fonts give much of the feeling to a site, and may even affect usability. For instance, English.SE benefits from using a serif font to clearly distinguish letters and make IPA more readable. Likewise, Judaism.SE's serif font makes Hebrew phrases (which are very common there) look much better. And on Math.SE the serif font makes MathJax formulas blend in nicely. […]

This was not a bug report. On Jun 7, Joe Friend ♦ posted status-completed.

@Monica Cellio, user and moderator on the following sites: Mi Yodeya, The Workplace, and Worldbuilding, on Sep 27–a month before the twitter debacle–posted

5. But we can change some of the design consequences, like fonts and accessibility and our logo. I appreciate that SE is trying to work with communities on the things we can affect, and upvoting the announcements and being constructive and just plain civil in our responses are both keys to that.
   After we responded to the preview of our design, Joe had this to say:

   You all get the nicest community award. Your feedback was on point and very constructive. You're all awesome. I've added tags to the feedback below and we will try to include changes where ever we can to address it. Thanks! - Joe
   And you know what? They fixed all the things we complained about on that post.

6. The Save the Robot! campaign was initiated on Worldbuilding meta on July 10th.

As per this post of the site's design being introduced, a lot of thought has gone into the site we know and love. It would be a great loss to lose these designs that have a place in our worldbuilding hearts.
Please upvote this to show your support in the current site-design, the true site-design.

7. On that same day, a similar request was posted by @Mithrandir on MSE

Second, what's happening with those figures? You've incorporated the header image, but... what about our friends in the theme there? Can't we fit them in somewhere?

The post on Workldbuilding attracted 83 upvotes with no downvotes while the post on MSE gained a respectable 74 upvotes. And the two heroic figures were restored on the new theme. I'd call it a success story.

I think there is enough evidence to suggest that meta is not without its merits but it seems that I misinterpreted the OP's protest because in the comments the OP spells out [emphasis mine],

The point of this post is for issues where the Powers That Be have not asked us for ideas or to tell them issues. Essentially, it's about their response to unsolicited advice.

But let's remind ourselves of the title. It asks

What are the effective communication channels for effecting change to SE?

Did the examples mentioned show that change can be affected through MSE or not? Moreover, SE staff did not ask if users liked the new theme, but to “provide constructive feedback” while SE's new community manager, Catija ♦, (a member of SE Meta for over 44 months), posted on every child meta site the following directive: “Please help us look for issues/bugs related to the theme design and how we have mapped the old theme to the new.” and “If you have concerns or issues regarding the left nav or the overall approach we are taking to theming, then this Meta Stack Exchange post is the right place for feedback.”

So, how accurate is the accusation that MSE staff barely (or never) respond to observations, complaints, or ideas on how to improve the Stack Exchange experience for its 170+ Q&A websites? If–as the OP argues–it's true that staff ignore users' suggestions and complaints and are only prompted into action when a problem is exposed in a tweet, then maybe, maybe, it's time for this white elephant to go into retirement. Tim Post's answer certainly seems to hint at this direction:

I think, in 2019, we're going to have to bite the bullet and replace at least bug tracking and feature requests with something else, or have something else behind the pipeline consuming meta and making sure it gets put somewhere else in ordered form so we can work from it.

We know this is broken, and it's something we need to fix. And part of that fix really is adjusting expectations - we just can't ship the stuff folks ask for nearly as fast as we could previously -- we're just too big and there are too many dependencies and stakeholders.

EDIT 31 October

I got a fair number of downvotes for this post but if you believe that the people who run Stack Exchange do not listen to advice, do not respond to ideas, and do not care about users' grievances, the elimination of MSM is the natural and only logical conclusion.

Do the downvoters agree that MSE is not a useful mechanism for pointing out problems? Do they sustain that users have no other venue but Twitter if they want to be heard or taken seriously? Or do the downvoters disagree that Meta should retire, that it's served its purpose? Frankly I don't know because (until a few minutes ago) the only comments I got were by the OP and by user1271772.

The essential question posed by the OP is a very simple one, is posting on MSM a waste of time? It seems that 147 users agree with his analysis. While another 64 users agree with Rand al'Thor's answer, which is scathing in its criticism and conclusion.

[emphasis not mine]

From a cynical point of view, a Twitter rant might put SE in a position where they have to do something about a problem. A meta post never will, because they control the platform.

---

ORIGINAL (and edited)

What could replace MSE?

Child Metas Fly the Coop

Each of the 170+ sister sites has its own meta site. Users are encouraged to post “a question about a specific site's function and policy, or a specific question on a specific site, … on the per-site meta.”

By eliminating MSE you place greater power and responsibility in the hands of child metas, and in the hands of the elected mods. Users who know and usually deeply care about the site. That can't be a bad thing.

With the advent of the new theme, and its adoption network-wide, every site will have the same identical platform, every site will have compatible themes, so every bug reported on any child meta should be reproduced easily on any of the other 170 sites. Any suggestions on improving a site's appearance or functionality will also have to take into account the other 170 Q&A sites. There can be no more requests to tweak the voting arrows on a specific site or modify the logo's position for instance. If a proposal on a specific child meta earns significant consensus the mods on that child meta can pass that proposal/petition to the other sites, when a minimum consensus is reached (maybe +200 upvotes) then the "powers that be" must post their response, that response can be posted simultaneously network-wide.

MSM can then be used for official announcements, games, competitions, and when the SE developers need feedback on new features and fixed bugs etc. Users can then posts answers, but ONLY answers. One of the knock-on effects would be the elimination of off-topic questions that continue to plague the site on a daily basis.

To sum up, it's all about numbers; community managers, product managers, software developers and the Director Of Community Strategy would listen to users if proposals were upvoted on several different child metas rather than posted on Facebook or in a tweet by a solitary user… unless their name happens to be Jon Skeet.

---

Kinda related:

Create a link between Meta from all the per site child meta sites

How to get attention to a post on a child meta site

Make network-wide FAQ posts available on per-site metas

Can you do a better job of informing new meta users of the MSE M-child difference?

How can we get people to participate on Meta sites?

Which meta am I supposed to post on?