Hi peeps and a very happy new year to anyone that sees this.
Below is the transcript of an idea posted to me by one of our top volunteers. It outlines a mechanic within Insided to prune inactive community member accounts.
Their reasoning is to protect against sleeper cell spam accounts who try to navigate the anti-spam detectors. I can also see a value in this with keeping an accurate idea of community member size.
It could be modified so that any account about to be deleted could get an automated PM to check if they object to this. That will help retention. Idea welcome!
I've just recently had another possible feature idea for inSided to consider. It's kinda related to anti-spam measures, but in some ways could also be useful for privacy as well.
This forum seems to have a huge number of sleeper accounts, either created by spammers or by legitimate users who created an account but never used it. I'm curious to get your thoughts on this as well.
What would you think of having the ability to automatically prune accounts that were created more than X time ago, but never actually did anything at all since creation (other than verify the account at most). The definition of sleeper account that I'd define would be:
Created more than three months ago
No custom avatar
No proper profile filled in
Zero topics started
Zero replies/comments
Zero Best Answers
Zero Ideas
Zero contributions
No other form of publicly visible activity (basically anything that isn't just visible in Control)
No PMs sent
Not currently banned? (optional?)
Any number of Like reactions
Any number of logins (simply logging in is not activity per se) or alternatively, fewer than 10 logins
NO rank and/or Newbie Rank
NOT a Super User, Community Moderator, Admin, Community Manager or any other kind of trusted user
Does NOT have any access to Control
Obviously, there's a risk that this could nuke accounts that have valuable contributions on, but I've considered that possibility. To avoid wiping out legitimate users, All of the above must be TRUE for the lifetime of the account for the criteria to be met. Pretty much any sign of life in that time should be enough to prove the account is not some sleeper account or a single purpose account. If any of those conditions are false, this rule should not trigger at all and the account should be excluded from being checked again.
The idea here is to basically try and pick up accounts that have clearly been either abandoned or might become spam zombies later on, but also trying to avoid affecting too many legitimate users. It should never pick up users who have contributed somehow at least once.
As for how to handle accounts picked up by this rule, I'd like to suggest some possible options:
Lock the account and if the user comes back, have some kind of unlock process
Put the first few posts through pre-moderation on the users return, then restore normal access
Restore the new user restrictions that would be otherwise enforced against new accounts
Allow the user to use the account "normally", but silently throw all posts into a moderation review for the first few posts, then release this state once the account is trusted
Assume the user doesn't want the account and delete the account completely, but not ban it - this is the most privacy friendly option for cases where the user doesn't want the account anyway, while also freeing up the username and allow them to re-register in the future, should they wish
It's a bit of a weird tactic, but I feel it might help to cut down on spammers abusing sleeper accounts and bypassing new account restrictions and improving user privacy in relation to abandoned accounts, while trying to maintain a reasonable balance for legitimate users who might just be on holiday, but are otherwise active.
Any thoughts on this?
