10

Is it just me, or do >50% of questions get instantly downvoted on SFSE? IMHO this is contributing to muddle the actual representation of the quality of individual questions, and even worse: demotivate new contributors, seeing their posts instantly downvoted.

I know that down- and upvoting is indeed quite subjective, and our standards and expectations differ a lot. That said, I believe only few of the posts downvoted deserved so. Seeing this level of consistency, also makes me think if this is something which is done by few, if not only a single member (I don't have the data to prove it).

How does this resonate with you? And in case this is done by one or few people, will it be sufficient to result in a warning or a ban?

Here are few examples, a.o. this, this, this, this.

EDIT 01/05-2024:

Really interesting discussion here, which obviously opens a much deeper problem than one or few people consistently downvoting questions (often shortly after creation, and without any commenting or attempts of improving them by editing). Am I to blame for having an itchy trigger finger, when it comes to downvoting? Potentially. Perhaps hanging around SFSE for nearly 8 years wears my patience thin. That said, I do try and comment with follow up questions, and instructions to new members, and close mostly obvious duplicates - downvoting only really poorly written, unsalvageable questions or answers provided by GenAI.

Due to the decline in questions, I have spent more and more time on Trailblazer Groups. But... I quickly realised how much I miss all the different tools in my kit, which SFSE offers. With my only ability to either upvote, or comment (by posting an answer... sigh), the community provides little means of providing critical feedback to askers. There also seems to be questions by people with slightly lower seniority on Trailblazer Groups, compared to here, where the technical level seems overall to be higher - even amongst the new members.

And speaking of members - it also seems that we are losing some of the more senior people here (at least in the area of SFMC), where the same 3-4 people are answering. 2-3 years ago, this was more equally distributed between at least 10 different names.

I am really worried this community is in a downwards spiral, and I will be very happy to contribute keeping it alive. Do I have the solutions and answers? Probably not, but I am always open to brainstorming and discussion us towards few potential scenarios for recovery.

6
  • That's an interesting observation. Is that for all types of questions or only for certain tags / topics? Also what time period, is this a recent trend? From my perspective, most of the time I see downvoted questions for pure spam, and occasionally see -1 or so
    – kurunve
    Commented Apr 30 at 21:08
  • Can you add some examples of where you think it's been unfair? I only typically view questions and answers that have been flagged, don't have much time for general controbution outside of being a mod!
    – Matt Lacey Mod
    Commented May 1 at 10:39
  • @MattLacey thanks for addressing this. I have few examples of downvotes, which are "in a grey area" of whether they should have been downvoted. They have been added to my question. I have a gutfeeling, that we are seeing some more agressive downvoting, possibly by one or few individuals. If you easily are able to identify whether this indeed is the case. IMHO downvotes, especially on questions by new contributors, should not be left without a comment. Commented May 1 at 12:00
  • I don't scan new questions all the time, but I feel like I've noticed this before when I do; I try to take a minute to consider if it is a good question and give it an upvote if so. Answers always get more votes than questions around here. Commented May 1 at 12:19
  • I often feel the same way about close votes. Some folks are really aggressive about closing questions. Commented May 1 at 12:28
  • Thanks for raising the problem. I rarely downvote, but regularly scan new questions and upvote what is lacking a comment and from my point of view unfairly downvoted. These days the majority of my upvotes are not in appreciation of a contribution, but in quiet rejection of equally quiet rejection. It feels like watering plants. (I dislike obvious bullshit and lazyness as much as anyone, and if anyone wants to downvote this - so be it! But if someone has taken some care in a post, my sympathy is with the OP.) Commented May 4 at 9:54

3 Answers 3

7

The "early downvote" percentage part is actually something that can be answered using SEDE (StackExchange Data Explorer) provided you know enough about Microsoft's T-SQL. I don't know enough about T-SQL nor SEDE to get a really good result here, but I did know/learn enough to toss this query together.

For the week of 2024-04-22, we had 136 total questions asked, with 61 of those being downvoted at least once within 5 minutes of being asked. ~45%

Going back to the start of this year (2024), that drops to ~27%.

Most (like, 90%) of the downvotes happen in the first 5 minutes, it seems.

Any sufficiently sized community is going to have differences in opinion on how to handle community issues. "When should a question be downvoted?" is definitely something that's going to have diverging opinions.

While downvotes are discouraging, and I do share the sentiment that more questions get downvoted (and close voted) more quickly than ideal, I don't think what we see rises to the level of being rampant/profound.

My personal thoughts:

  • While Stackexchange does heavily encourage participation by voting, not providing a vote is itself still a vote of sorts
  • I feel like we get a large number of poorly asked questions (unclear, little/no research, mostly just a code dump, etc...)
  • While I think I tend to not downvote things, seeing a downvote or two doesn't seem excessively negative. If it gets to -3 or lower, then it could be an indicator of "dogpiling"
  • If I think a downvote is sufficiently unjustified, I'll upvote it to offset
  • I do leave a lot of comments, but I will skip doing that if I get the feeling that it'll fall on deaf ears (why put in the effort if it yields no results?) or if I think that it's sufficiently obvious (code dump with no explanation, repeated "do my work for me" questions, yes/no questions that are answered by the first google result)
  • Leaving a comment does, to a degree, paint a target on your back

My opinion for the 4 examples provided so far (at time of writing):

  • #1 I don't know enough about Marketing Cloud, doesn't rise to the level of vote-to-offset
  • #2 Mostly a code dump, user hasn't responded to my comment, -1 is probably fine, I didn't downvote this
  • #3 Possibly deserving a vote-to-offset
  • #4 Seems like a basic issue and a clear error, no evidence of research. I would've considered downvoting if OP needed more hand-holding than that. -1 is appropriate here, I think

I think that publicly providing a query that could narrow this down to specific users is a bad idea. If people think this is a sufficiently large problem, I can see if I can get a SEDE query put together and reach out to one of our mods with the results.

+edit: While SEDE does say that votes have a user id attached to them, it's always just null for me (even logged in). So we can't use that to get an idea of whether this is due to a small group of users or not.

2
  • From my perspective, things are pretty shocking.
    – sfdcfox Mod
    Commented May 1 at 16:39
  • Your flags don't go unnoticed!
    – Matt Lacey Mod
    Commented May 1 at 20:40
7

It's not just you. We have a systemic network problem, which I don't think anyone has bothered to look at for some time. Unsurprising, since we don't have many people who can, and we're volunteers. This network's positive experience feedback metrics peaked about seven years ago. Since then, there has been a steady decline in the number of questions, answers, and upvotes.

Our other metrics, like visitors, new users, and activity, are still fine; they're on par with our metrics from seven years ago. Unfortunately, our downvotes, closes, and deletions are also on par with seven years ago. They haven't seen a relative decline compared the other positive feedback metrics, but instead remain fairly level.

This means that over the past seven years, users are relatively more likely to experience negative feedback than positive feedback. Back then, it was something like 90% positive feedback, while today, it is indeed closer to 50%.

I'd have to additional data mining to try and figure out how many of those affect new users, but based on the reputation leaderboard, I would hazard a guess that those at the top receive the majority of the positive feedback, and the majority of negative feedback would be directed at new users.

Changing to the voting leaderboard, those at the top are more likely to have engaged in more negative feedback than those lower down. The top 180 users probably have close to 99% of the votes, out of approximately 3500 users. I personally have approximately the same number of votes as all 36 users on page 5, to put that in perspective. However, I have less than 300 lifetime downvotes, or less one a week on average.

So, while it is a small number of users that are consistently providing negative feedback, I don't think it would be appropriate to ban some of our biggest contributors. I won't call out names, as these people know who they are (and this information is public anyways), but I would like to say something important.

We Are Ohana

To quote one of my favorite movies:

Ohana means family, and family means nobody is left behind or forgotten.

We are part of the Salesforce Ohana--the family knit together by a common ecosystem. We take care of each other. We welcome each other with open arms. We help those that need help. This is what it means, to me, and I like to think my own metrics are a reflection of that commitment, rather than just some numbers to keep cranking up.

We should all strive to embody that welcoming attitude. Let's add constructive comments for more information or clarification instead of downvoting. Let's edit a new users' post for formatting and professionalism, rather than just downvoting or closing as off-topic or whatever. Let's upvote more questions that are even remotely positive. It is up to each of us to make new users feel welcome.


Edit: I also just noticed the examples. Here's my take on them.

Using outbound template for reserved keyword 'yes' and Y

This is fine, perhaps a minor edit for grammar and/or punctuation. Someone with Marketing Cloud should be able to answer, or it could be researched.

Send Email through External Email Services by using Apex and VF Page

I'd comment that this should be trimmed down to just the necessary code, and maybe having them check the Email Logs, and check spam folders. I don't see a need for a downvote here. They have a legitimate question that should have a legitimate answer.

How to get current user related data from apex class to an aura component?

This one is fine, too. The answer is slightly more complicated, though, which is probably why I missed it.

Getting error while implementing the trigger dispatcher architecture

Another decent question. In fact, Derek even answered it, which means it probably should be at least neutral. With an edit, it'd probably even be worth a few upvotes.


These are a great set of examples. These questions may have been somewhat questionable, but I believe minor edits could save all of these. This probably warrants a conversation with the core community members (and anyone else that wants to join) about how we can improve our reception.

6
  • Wow... ok. I didn't know what Pandora's box I was about to open. Comparing yours and Derek's response only supports my point of voting being highly subjective. I do agree to your points on the overall decline in volume and quality, and I have also seen these stats decline for some time now - as this discussion (and Adam's query linked in his answer) proves, "our" tag peaked in 2017, and its decline accelerated in 2020. Commented May 1 at 19:25
  • @LukasLunow Voting is subjective, true enough. But my answer is much more focused on the state of affairs. We must address a problem that will not right itself without intervention. If we wait too long... I think you get the picture.
    – sfdcfox Mod
    Commented May 1 at 20:09
  • 1
    @LukasLunow one thing that would help is to use the tools available to help counteract the downvote. You could upvote questions that you think have been wrongfully downvoted (not great I admit), or edit questions where grammar whatever is slightly flakey. Mods can't see who's voted which way on a question, so it's tough to know if we have serial downvoters out there.
    – Matt Lacey Mod
    Commented May 1 at 20:39
  • 1
    @MattLacey True, but we can look at the voting leaderboard, and see each user's public stats. Some users have about a ~2:1 upvote/downvote ratios, which I feel is overly aggressive, compared to my ~72:1 ratio, which is probably not aggressive enough... If I had any say in this, I'd suggest we shouldn't have more than maybe ~10:1. We're ever under-upvoting, over-downvoting, or both. I'd like to see that change going into the future.
    – sfdcfox Mod
    Commented May 2 at 2:21
  • 3
    As I alluded to before I tend not to look at questions so much on the site these days - I just deal with flags and the like, but I might make a conscious effort to review new questions and upvote those that are well written. Maybe if more of us do so it'll help allevieate any perceived issues.
    – Matt Lacey Mod
    Commented May 2 at 9:45
  • 1
    I think it would be great if we could mandate a comment is required for a downvote, I've noticed this downvote behavior and while some are obviously for poor/no effort questions often I am left wondering with the why.
    – Girbot
    Commented May 8 at 10:31
4

Perspective: I am not a new user to SE but I am a new user/salesforce developer. 25 year programming vet in different languages and platforms. I've worn all the hats and truly love modern programming. I'm currently blending my own pre-salesforce experience with a new SF centric way of doing things. I do rely on SFSE more than trailblazer as I prefer the interface here, but I will let Google take me to trailblazer for some other answers.

I've also posted questions that were downvoted to 0 without a comment. @sfdcfox still went ahead and answered my questions which was great, but then we have a 0 vote question with a high-quality answer. I've found this curious enough to take time out of my day to come meta about this!

Here's an example question that was downvoted to 0 when I stumbled onto it. This question is highly pertinent to something I am working on and so I left an upvote on it.

I would like to understand why a question like this was downvoted. Is it possible to have no question downvotes without comments?

P.S. Thanks to the mods here.

Edit: Here's a question I just posted and it was immediately (within 1 second) downvoted without comment.

1
  • 1
    Unfortunately, we don't control the network to the level where we could require a comment for a downvote. This is up to Stack Overflow to implement, should they choose to do so. I personally think it'd be awesome if we could make it so that one must either leave a comment or upvote a comment on the question before the vote sticks, but unfortunately, that only applies to Votes to Close (VTC). Even that model is flawed, because once closed, no answers can be given, and I've had that happen to me a few times, typing up an answer only to be told I can't post.
    – sfdcfox Mod
    Commented May 8 at 13:35

You must log in to answer this question.

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