The difference is that I’ve only ever received notifications from automod as a reaction to something I did, and those notifications came in either as a comment reply or as a direct message.
I looked it up - the last time I interacted with this community was 5 months ago. I’m not a subscriber, either.
I can understand the importance of casting a wide net for awareness, but I think your scoping and your approach was wrong.
If I may offer a few suggestions - I don’t think the idea is wrong, but I think that the messages should have been direct messages to users that clearly explained why that individual user was receiving the message, plus whatever else you needed/wanted to say.
I would have had more rigid criterion for inclusion, too. Subscribers, certainly, and perhaps people who have interacted with the community a minimum number of times within a limited time period. For me, 3 times in 3 months seems generous enough to cast that wide net while also not being overly-broad.
So if I were executing this, I would have had the bot direct message people, and say “Hello, I’m a bot acting on the behalf of X community. You are receiving this message because you are a subscriber or have recently interacted with the community. We wanted to let you know that the community is moving! It can be found at Y.”
The bulk tagging action was, I think, especially egregious. Posts with massive amounts of tags are a very common spam technique and people automatically have a negative association with them. Pair that with non-specific criteria and relatively non-descriptive explanations, and I think it’s a perfectly reasonable thing for folks to have adverse reactions.
I hope that you consider the above as fair, and not overly critical. I’m not trying to throw stones. Everyone is human here. Well, except for the bot. But even it makes mistakes.
Thank you for your nice feedback. I just created a thread on !fedigrow@lemm.ee , I’ll definitely include this.
FYI, at the moment the only people who can know the list of subscribers to a community are that community’s instance admins, via the database directly, so not something very accessible for mods.
I had not considered that as a possibility/did not know.
That definitely throws a wrench into the subscriber only idea - at least not without admin support.
I assume that most instances have some rule against vote manipulation. A complementary idea I had was to - for a set period, make a post restating the change and request that community members upvote it for visibility. Such as 1x post a week or what-have-you - varying the day to accommodate for folks on different schedules. (Workday folks, vs weekend folks, etc)
Definitely an option without the request for upvotes, regardless of instance rules, but perhaps without as much reach.
I hope that you consider the above as fair, and not overly critical. I’m not trying to throw stones. Everyone is human here. Well, except for the bot. But even it makes mistakes.
Not at all, I’d say all well reasoned feedback is always welcome, and I mostly agree with all your points.
To mention a things:
The mass tagging instead of DMs were to spare the server from spamming DM requests. Originally though, the bot was written to reply to the latest post or comment you made in the community, but similar to sending DMs, mass tagging in a few condensed comments is “tidier” in some sense.
The script (it isn’t really a bot as it requires manual input) was set to look 365 days back, which might have been fine for football@lemmy.world where it was originally used. I suppose for a community that have more non-subscriber engagement, a tighter timeline like a week or month *might *have been better.
If I could have gotten the subscriber list via the API, I would have done so. I suppose a coordination with the admins to get the subscriber list would be better.
The difference is that I’ve only ever received notifications from automod as a reaction to something I did, and those notifications came in either as a comment reply or as a direct message.
I looked it up - the last time I interacted with this community was 5 months ago. I’m not a subscriber, either.
I can understand the importance of casting a wide net for awareness, but I think your scoping and your approach was wrong.
If I may offer a few suggestions - I don’t think the idea is wrong, but I think that the messages should have been direct messages to users that clearly explained why that individual user was receiving the message, plus whatever else you needed/wanted to say.
I would have had more rigid criterion for inclusion, too. Subscribers, certainly, and perhaps people who have interacted with the community a minimum number of times within a limited time period. For me, 3 times in 3 months seems generous enough to cast that wide net while also not being overly-broad.
So if I were executing this, I would have had the bot direct message people, and say “Hello, I’m a bot acting on the behalf of X community. You are receiving this message because you are a subscriber or have recently interacted with the community. We wanted to let you know that the community is moving! It can be found at Y.”
The bulk tagging action was, I think, especially egregious. Posts with massive amounts of tags are a very common spam technique and people automatically have a negative association with them. Pair that with non-specific criteria and relatively non-descriptive explanations, and I think it’s a perfectly reasonable thing for folks to have adverse reactions.
I hope that you consider the above as fair, and not overly critical. I’m not trying to throw stones. Everyone is human here. Well, except for the bot. But even it makes mistakes.
Hello,
Thank you for your nice feedback. I just created a thread on !fedigrow@lemm.ee , I’ll definitely include this.
FYI, at the moment the only people who can know the list of subscribers to a community are that community’s instance admins, via the database directly, so not something very accessible for mods.
I had not considered that as a possibility/did not know.
That definitely throws a wrench into the subscriber only idea - at least not without admin support.
I assume that most instances have some rule against vote manipulation. A complementary idea I had was to - for a set period, make a post restating the change and request that community members upvote it for visibility. Such as 1x post a week or what-have-you - varying the day to accommodate for folks on different schedules. (Workday folks, vs weekend folks, etc)
Definitely an option without the request for upvotes, regardless of instance rules, but perhaps without as much reach.
Yes, otherwise we would have probably gone that route.
Your other idea would probably be seen as vote manipulation indeed unfortunately
Not at all, I’d say all well reasoned feedback is always welcome, and I mostly agree with all your points.
To mention a things:
The mass tagging instead of DMs were to spare the server from spamming DM requests. Originally though, the bot was written to reply to the latest post or comment you made in the community, but similar to sending DMs, mass tagging in a few condensed comments is “tidier” in some sense.
The script (it isn’t really a bot as it requires manual input) was set to look 365 days back, which might have been fine for football@lemmy.world where it was originally used. I suppose for a community that have more non-subscriber engagement, a tighter timeline like a week or month *might *have been better.
If I could have gotten the subscriber list via the API, I would have done so. I suppose a coordination with the admins to get the subscriber list would be better.