46
TL;DR;

Our overall goal is to communicate early and often, especially if we know that some of our commitments are changing. With this post we want to share that delivery dates for some Q3 projects will be shifted to Q4.


When we move into a new house everything goes smoothly for a while. But one day the entrance light bulb burns out. Not a problem! We’ll fix it the following Saturday morning. Another day the tap in the kitchen stops working. That is fine as well! If we cannot fix it this week, we will definitely do it next weekend. As luck would have it, when the next week arrives we have to do some gardening in addition to the tap, and there is some water in the attic out of nowhere…

I think we are all familiar with the situation where the number of support / maintenance tasks starts to snowball and “Saturday morning” becomes insufficient. Usually, there are two choices at this point, we either get used to living with a leaky roof or we take a few days off of work to fix all the issues at home.

No tomorrows. No excuses. Ticket Smash on September 8th - September 18th

In total we have almost 600 unhandled CM escalations today. We do apologize for that. To remedy the situation, we have decided that the entire Community Team (Sara Chipps, Tim Post, Stephanie, Juan M, Catija, Cesar M, JNat and Nicolas Chabanovsky) will be exclusively working on CM escalation requests for two weeks, from September 8th to September 18th.

What is a CM escalation?

Each Stack Exchange site has its own superheroes (we call them “moderators”), the users who help the community handle exceptional situations. Sometimes the existing moderator tools are not enough, and a moderation action should be performed manually. For example, to investigate a complicated voting fraud we need to query the database directly. In such rare cases, moderators send the CM Team an escalation request and CMs take further actions in regards to the issue.

The “Few Days Off”

When we started experiencing difficulties with handling tickets in the first place back in 2015 we hired two CMs to work on tickets exclusively. The house was warm and dry until August 2019, when the roof started leaking, i.e. the response time began to grow little by little up to today, when we have 600 tickets in the queue.

We either go to work or fix the roof at home. Users have always been, are and will be a priority for us. As a result some of the projects expected to be delivered in Q3 will be shifted to Q4 because of the current situation with the escalation request queue. We will continue working on improving our platform as soon as the escalation request queue decreases.

All the delayed projects will be on our Q4 roadmap that we will publicly share with you in October!

Update

I posted a wrap up of the event as an answer to another question.

4
  • 13
    I admire you for the job you do, but huge thumbs down for the company management that cut the CM team drastically and led you all to the pit you're now in. And I'm sad and disappointed they do nothing to remedy this. I won't quit or rage over this, just sad for you and the rest of the CM team. Commented Sep 11, 2020 at 11:35
  • 4
    Thanks for taking this effort. Just got some serial downvoting reversed. If you are handling things chrnologically you might be nearly done, as that flag was from last july.
    – Luuklag
    Commented Sep 11, 2020 at 12:34
  • 4
    It's been a 11 days since the CM team are working only on this. How is the queue now?
    – llrs
    Commented Sep 29, 2020 at 9:49
  • 1
    Related: CM escalations - How we got the queue back down to zero (posted September 30, 2021)
    – V2Blast
    Commented Oct 4, 2021 at 18:48

5 Answers 5

39

Good to hear, and thanks for letting us know - the transparency is appreciated.

I've got a sneaking suspicion that one reason things have built up so much is a loss of both, well, people available to look at the queue - you've lost three people from the CM team who have yet to be replaced - and a loss of institutional knowledge; Shog especially, if I recall correctly, would handle all the difficult voting fraud cases that were outside animuson's ability to handle.

Hopefully y'all have some internal documentation about various ways to look into this, and maybe some magic scripts laying around that nobody fully understands.


Incidentally, the CEO had mentioned a commitment to hiring new CMs, to replace the CMs you've lost (and if we're really lucky even an expansion to the team). Is there any update on when that's going to happen? That'll probably help prevent this from happening again in the future, in any case.

5
  • 26
    FWIW, I had a backlog of vote anomaly tickets going back to 2017, so ain't like I was keeping up with that either... ☺️
    – Shog9
    Commented Sep 8, 2020 at 18:58
  • 7
    Leaving the fun jobs for others, how generous of you @Shog9
    – Luuklag
    Commented Sep 8, 2020 at 19:07
  • 16
    Yeah... The problem with a lot of those tickets was they're concerning problems that are either very hard to see or very hard to fix - so it wasn't unusual to have to write new scripts (or modify existing ones) as part of handling them. Over time, that meant the tools got better - but it did tend to delay things.
    – Shog9
    Commented Sep 8, 2020 at 19:11
  • @Shog9 yeah, cookie jar an all that.
    – Braiam
    Commented Sep 8, 2020 at 19:45
  • 2
    @Shog9 I do wonder what caused this backlog...
    – Magisch
    Commented Sep 10, 2020 at 13:31
23

I'm glad that there's recognition of the issue - especially as SE works towards trying to work through improving their relationship with the community. It's about time.

It's not as impressive seeming as bigger projects but the day to day mechanics of community work and the understanding that our escalations will be handled is essential for the volunteer mod corps to function.

I do suspect though that the genesis of the backlog probably goes further back than August 2019.

Once cleared, I hope it's kept in check as opposed to spring cleaned when it looks really really bad.

I'm wondering if the resolution of CM tickets should be considered as part of the company's commitment to the community. These escalations do for a significant extent, are for issues we are unable to deal with as mods.

Formally or otherwise, it seems like a good indicator of the level of resourcing and how well the company is doing in that respect. It would be nice, once the backlog is cleared, to consider and share the state of the queue and relevant status statistics for tickets much like any other indicator of SE's commitments towards the community.

2
  • 12
    At one point we were sharing statistics about the size of the queue with the moderators in the Teachers' Lounge every Friday. I don't know whatever happened to that (it was JNat's domain). Maybe he could chime in. I believe most moderators found the information useful.
    – animuson StaffMod
    Commented Sep 8, 2020 at 21:52
  • 9
    Oh, wow, that was a while back, I'd forgotten I was doing that. Ended up dropping it when I stopped working on our ticketing system, but I'll bring it up with the team since there's prolly some benefit in bringing that back
    – JNat StaffMod
    Commented Sep 11, 2020 at 11:35
18
+50

I wish I could award a bounty to your question. But here is an answer instead.

Thank you!

Thank you for taking care of us, and about informing us in advance. I hope you guys succeed in smashing that entire queue in such a short amount of time!

13

As Journeyman Geek mentions in his answer:

I do suspect though that the genesis of the backlog probably goes further back than August 2019.

Whew, that's going back quite a ways if that's true.

So then... What sorts of changes are you looking at to improve response time to these escalation requests, if any?

I know that that's a very vague, open-ended question, but I'm sure that you and I both see the same problem here. That kind of backlog is pretty... rough - and leaving tickets unresolved like that for extended periods of time can be a bit frustrating for all parties, I imagine. I'm curious what steps y'all are going to take to ensure that response time is minimized as much as possible.

8

This is good to hear - thank you, and thanks for telling us. I have confidence that the Community Team's efforts will be more than enough to obliterate such a large queue in that small an amount of time. Sara, Tim, Nicolas, JNat, all - good luck.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .