Stack Exchange's growing pains
Nov. 20th, 2018 10:01 pmStack Exchange used to be able to function like a smaller company...until they couldn't. They don't seem to know how to be a bigger company yet, so sometimes they step in it badly. This time they not only stepped in it badly but they then reached for the shovel to dig even deeper.
Background: If you visit any site on the network you'll see, partway down the right column, a list of random-seeming questions from other network sites. These are called "hot network questions", and the communities have been asking for years for SE to tune the algorithm that chooses questions. (It responds to velocity, not quality, and thus optimizes for controversy.) People complained; nothing happened.
In mid-October somebody who turned out to be a troll complained on Twitter about two such questions, seen on Stack Overflow, from the site Interpersonal Skills. The title of one of them was not great (which is what edits are for); the other one was fine. But this person got a rant on and has followers. Within 40 minutes, an employee responded with something like "that's not ok; I've just removed that site from the hot list and we'll look into what's going on with that site". Great way to throw a community under the bus there. (The community wasn't notified until hours later.) Meanwhile, one of the moderators on that site, who I know to be a clueful and thoughtful person, responded to the tweet (in retrospect a bad idea) and tried to help. Other people responded too because, hey, that's how Twitter works.
So then our Twitter troll (twoll?) ranted some more because people were responding, and accused the moderator and others by name of "sealioning" (apparently a form of trolling) and generally spouted outrage, and a different employee jumped in and said something like "if those messages came from mods we'll fire them" -- without even asking first what these allegedly-trolling messages said. (The employee thought they were direct messages, meaning you'd have to ask because DMs are private.) So the employee jumped to a faulty conclusion and validated the troll without seeming to consider that maybe the facts were not as presented.
Stack Exchange is way the hell overdue for some internal education on how to do social media. They admit that, but meanwhile they have employees who've helped to malign named volunteer moderators and one entire community, and we moderators kind of thought they should do something about that. Like, maybe, apologize and say it won't happen again. Or at least retract the tweets. Or something. (Employees are allowed to have private opinions, but when you speak as an employee you need to be more careful.)
The next day Tim, a manager on the community team, posted on SE's meta site asking for ideas about how to change hot network questions. In that post he said "Some things happened yesterday that caused a need for us to (quickly) remove a site's eligibility to contribute to the list of hot network questions." "Need"? Even after the initial round of discussion, this wasn't being treated as an over-reaction. Sometime in the next few days the affected community was told that this hasty, troll-induced change would not be reversed any time soon.
The Twitter-storm started on a Wednesday. We were promised a blog post supporting moderators and clarifying policies "soon". With no motion after several days I wrote Dear Stack Overflow, we need to talk on Medium, and to get it seen I tweeted it.
We’ve had a rough few days. I get that you’re tired of hearing about it, but the damage is still there, so we can’t just ignore it, hide behind the weekend, and hope it’ll blow over. It won’t. You need to act.
Your silence in the face of bad behavior is harming your relationship with the volunteers and community members who make your sites work.
I spent almost as much time on the cartoon as on the post, by the way, because (a) I suck at graphics so it takes a while and (b) I knew both Medium and Twitter would use a graphic in their previews if present. You should click through and see my glorious art (cough).
That post caused a ruckus around the network; I saw lots of comments along the lines of "if Monica is mad enough to write that, it's serious". Another moderator made this meta post in direct response to my post, in which he wrote:
We've been really really patient, but we shouldn't need to tweet at folks to get your attention. We've been really patient. We really would love SE to grow, but not at the expense of its heart.
Tim and Jay (VP on the community team) both tweeted to say that they value our moderators, which was a good start, but it really isn't enough.
Almost a week after my post, the moderator who had been personally attacked on Twitter wrote a post of his own and tweeted it. It's a very thoughtful piece, and it generated a reply on Twitter from Jay (there's a thread there). It's a thoughtful reply but, like the earlier one from Tim, seemed to make excuses for the employees instead of apologizing for their blunders. Nobody has yet apologized officially. Team members refused to even ask the employees to clarify that they weren't speaking for the company, saying that would be throwing employees under the bus (not true).
A few days after my post (and before the other moderator's post), Jon Ericson, a community manager, made a post responding to mine. He's the only one I've seen so far who gets that SE broke trust with its moderators and communities. He went on to explain how SE can't operate the way it did when it was smaller, and he's optimistic about some of the changes they're making. I appreciate the (personal, not company-representing) response, especially coming from somebody I've locked horns with in the past, but it's necessarily incomplete.
That was all in October. The promised blog post did not appear the following Monday, or any other day that week. (Then bad things happened here in Pittsburgh and I wasn't as focused on this for a while.) Nearly five weeks later that blog post has still not appeared. It's ridiculous at this point. I don't think we're getting it (and they don't want to tell us), and I also think they've squandered the chance to use it to repair the damage they did. As I wrote in a comment on Jon's post two weeks ago:
I was expecting a blog post a few days after the incident with some navel-gazing, some after-action review, and some clear statements (like Jay's tweets) on our own site. I fear that the expected benefit from that blog post is dropping off as the weeks go by, and that too makes me sad and frustrated. I never expected weeks to go by like this. And I know this is wearing on members of your team who were already stretched too thin before this happened, so I expect there's some resentment on your team (not singling anybody out), and that doesn't feel so good either. Some of today's TL discussions were, um, not good in that regard.
(TL is the Teachers' Lounge, the moderators' private chat room. A senior CM had basically told us to shut up about it.)
So, to recap:
- A troll got SE's attention and SE responded carelessly.
- Having had the problem pointed out, they chose to retract nothing. SE totally failed at damage-control.
- They also chose to not reverse the hasty decision about that site pending a more thoughtful discussion about what should really be done.
- They did tweet to say our mods are valued.
- Nobody has apologized publicly on behalf of SE.
- SE has not publicly clarified its social-media policies.
- Near as I can tell, they're just waiting for it to blow over and expecting business as usual from the volunteers.
I'm feeling pretty demoralized. I love some of my communities and I'm not going to let SE spoil them, but I'll admit that on the community that's the most challenging to moderate, I'm not doing much right now and I'll decide later whether to resign. Another moderator on that site quit -- not just because of this, but this contributed. Several other moderators (not just on my sites) are visibly unhappy, though we are but a small proportion of all moderators. SE can probably afford to alienate us.
I'm also feeling like I dodged a bullet when I didn't go to work there earlier this year. SE is changing from the company that had attention-span for and good collaboration with its volunteers, and I can't tell if they know how to actually manage the transition to whatever comes next.
Edited to add: I said this elsewhere and want to say it here too. I believe that all of the individuals involved have good intentions, are trying to do the right thing, and got in over their heads. Stack Exchange corporately seems unable to fix this, and I think that's at least in part because they are no longer a small, well-functioning company. I think there must be a fair bit of dysfunction in place structurally.
I wrote this post mainly to collect all the pieces in one place, because this has been happening on Medium and Twitter and Jon's blog and Meta SE. But I also wrote it because this (not Medium) is my journal, a place where I write about SE and many other things, so it belongs here.
(no subject)
Date: 2018-11-22 06:26 am (UTC)(no subject)
Date: 2018-11-22 06:00 pm (UTC)We have sites about gaming (more on that in a bit), politics, skepticism (validating or disproving claims, which are often sensational), lifehacks, puzzles (where question titles are part of the puzzle and, thus, all over the place), worldbuilding, and more. A few years ago some hot new video game came out (don't know which; not into that world) and we started seeing first-person questions about murder and rape -- and that wasn't enough to fix HNQ (or remove that site). We see questions sometimes from the site for role-playing games that raise eyebrows (RPG doesn't just mean wizards and elves). We have seen questions from Workplace about sexual harassment and racial discrimination, and one I remember from a manager who wanted to know how to not hire women on his team because of religious issues (super-modest team members), and one from a Muslim asking how to not shake women's hands. We have seen questions from both Christianity and Islam about why God hates Jews. (Also homosexuals, if I recall correctly.) I don't think we want any of that on the Stack Overflow front page.
The algorithm that chooses questions is fundamentally flawed, sites have no control over what goes on there short of withdrawing (valid!) questions entirely, and it's caused problems before. And now, when a problem rises to the level of "we have to do something", they play whack-a-mole instead of removing the hot list from SO until they can fix it. They're just asking for the same issue to come up next week with that controversial question from Politics about illegal immigrants, or the one from Medical Science about erectile dysfunction, or the one from Worldbuilding about switching a creature's gender (and it turns out to be a question about aliens that change based on environmental factors).
Also, since it looks like you think it's obvious but it's not to me, what is sexist about any of those question titles? Maybe if I saw that the reaction would make more sense. I see a question about discouraging unwanted sexual advances -- actually, I see two like that. I get how the autism one looks bad without the context, though to me it's not "ableist" (obviously others disagree).
As for the moderator who responded, he was trying to help and got it wrong, but was accused (validated by SE) of "sea-lioning", which ascribes malicious motive. No SE employee ever refined that implicit accusation. Nobody stood up and said "we're sorry, our ill-trained moderator was acting in good faith and we should have provided guidance". Nothing, just accused of trolling. I know the mod, work with him on one of my sites, and I am confident that there was nothing malicious in anything he wrote.
ETA: Graham shared his response in Teacher's Lounge for feedback before posting it. Nobody said anything like "hey, better not". (I don't know if any employees were present at the time.) Whatever clues were missed were missed by a bunch of people, not just the one mod, hence "ill-trained" in the previous paragraph.
(no subject)
Date: 2018-11-26 09:39 pm (UTC)Graham tweeted 4 tweets in 2 minutes to "correct" her misunderstanding of the site. That reads as aggression on Twitter. Graham has tweeted 200 times and the person who complained has tweeted > 70k times. Probably, Graham doesn't understand Twitter norms. That's SE's problem. They tried to address it (but poorly and without fully understanding the interaction; agreed there).
Graham acted in good faith, but he made the situation worse.
The SE employee who responded also acted in good faith, and ALSO made the situation worse.
Those titles are offensive because they are interjecting sex into someone's workspace. Full stop, that's the reason. Sex is an inappropriate topic at work, and SO is primarily used for work (and school).
And in tech, sex is so prevalent! It's exhausting to constantly be watching out for SURPRISE SEX TALK, so seeing it on SO and just closing your browser window and never going there again is a completely understandable experience to me.
I agree; the best reaction would have been to remove the hot questions list rather than to remove a single site from it. Perhaps they will. Perhaps we'll have to keep pointing out how stripping questions of their context alienates people from SE.
But to be honest, to see AN ACTION TAKEN in response to a complaint about sexism at a tech company (even if it wasn't the best possible action) was such a relief to me. Thank goodness a company in the tech industry (that I donated my labor to!) takes this stuff seriously!