Yeah this is where the subjectivity comes into play! I consider mass following to be engagement hacking too, not just follow + unfollow.
bsky.app/profile/enga...
@engagement-hacks
Labels accounts suspected of engagement hacking (e.g. mass following accounts in an effort to get return follows, especially if unfollowing afterwards). Mostly manual. DM (or @, or reply) to appeal.
Yeah this is where the subjectivity comes into play! I consider mass following to be engagement hacking too, not just follow + unfollow.
bsky.app/profile/enga...
Yes, I consider following a lot of people you don't engage with to be growth hacking too.
Agreed that this is where it gets a bit subjective what everyone considers engagement hacking vs not! bsky.app/profile/enga...
Heβs following 15k people and not engaging with any of them - thatβs exactly the point of the label.
Did he not talk to any of you on Twitter either?
cc @bnewbold.net I know you've been thinking about labelers lately.
I've been wanting to write this thread for like a month, and it's been tough. Thanks to all of you for any feedback & thoughts.
Just shut down: I am very reluctant to do this, since I think it provides a massive user benefit to having around (& again, am sorry that I havenβt been able to fully keep up with the reports). The costs for hosting are minimal ($15 a month or so) so will probably not seriously consider this.
(4/4) I also wonder if creating more specific labels (e.g. βgofundme mass followerβ, βfollow for followβ, βpromoting somethingβ etc) would help with usefulness for people who differ on what they consider βhackingβ - please let me know your thoughts.
(3/4) Things like: how many people are they following, are they interacting with people (especially people *not* just replying to their own posts), are they promoting things [links] in their posts and/or descriptions, are they following bot / ghost accounts that donβt post anything, etc.
(2/4) But latelyβ¦ Iβve come around to the idea, in part because it would make me have to more clearly define what characteristics I help use to evaluate reports.
Add additional volunteers (1/4): This is probably something I should have started doing post-election in general, once we saw an influx of users here on Bluesky. I think that I was reluctant to do so because of how subjective the calls of βengagement hackingβ are.
(2/2) I think Iβd rather figure out how to export existing labels and offer that up as a bootstrapping option for a new replacement label and keep this around for a bit in maintenance mode.
Transfer ownership to someone else (1/2): Very possible, but Iβm quite concerned about turning over an existing pre-running labeler to a new owner who could use it maliciously.
So: what next? I dunno. Iβm very open to ideas from the community! Some things that have been floating around my head:
1. Transfer ownership
2. Add volunteers
3. Shut down (unlikely)
The other issue is, I think, peopleβs expectations of what constitutes βengagement hackingβ and what doesnβt. They arenβt all the same! And this makes it hard to meet the subjective guidelines of people.
Ozone (the mod tooling) does allow me to mute reporters, but I have been pretty reticent about doing that for users permanently. At this point, itβs like half the reports or more.
One of the major issues is people reporting non engagement hacks to this labeler. Depending on whether they include information in the report or not, it can be very obvious (βimpersonating Xβ) or not very obvious at all, which wastes time.
To be clear: I donβt do this for money (although Iβd love if Bluesky had its own hosting platform that it offered labelers for free so I wasnβt paying for my hosting costs).
Labelers are volunteer operations, first and foremost. At the same time, I realize how valuable they can be (hell, I started this *because* of how valuable I found the old Aegis label for engagement hacking).
Itβs also led to a massive influx of new users who are fleeing X (welcome!) and bringing some bad behaviors along with them (please donβt).
Why the burnout?
Obviously - national politics in the US. I cannot do a Twitter 2016 repeat of just mainlining every thing the current government seems to want to break, and thatβs led to me reducing the time I spend on Bluesky in general as itβs what most people are (reasonably) posting about.
Iβve had to add some automation to keep up with the influx of reports, but anything not triggered by that is still manually reviewed by me, which takes time and energy to seriously evaluate if the label should be applied or not.
Iβm still trying to handle the most egregious reports, but the overall queue (we get like *50* reports a day) is still growing. Even if I wasnβt burned out, this would still be a massive amount of work every day to keep up with.
Labeler Update... Unfortunately, not a happy one. Frankly, Iβve been burned out bad since the holidays and havenβt really gotten much motivation.
"Because he follows a lot of people" Yes, generally that's the case.
You are welcome to unsubscribe if you don't care about how many users someone is following.
He's following 15k users but does not do anything besides post photos.
I think Brooke is a good example of whitelisting, I don't think Nancy who doesn't any sort of engagement is a good exception.
I have been meaning to write up a thread on how I've been burned out re this labeler & some ideas, but the burnout's been holding me up. Always appreciate the feedback!
Appreciate it! Label has been removed, let me know if you have any other issues.
FWIW I do really wish (re my personal account, and as a label operator) that explicit following would override labelers so if there's a couple accounts labeled, they'd just get shown anyways. Or at the very least flagged so I can choose to keep following them or follow the labeler.
I think you can select "Warn" instead of "Hide" if you go to the labeler profile (aka @engagement-hacks.bsky.social) which should show that it's been hidden instead of making it non-existent?
Why does it follow 17k people if it just posts a Nancy comic a day? (I say this as a Nancy fan)