I keep close tabs on the ‘suspect’ user category and clean this up on a weekly basis. After doing some digging in data explorer though, I found a LOT of spam accounts that apparently did not match the suspect criteria. What they all have in common is that they have super low to non-existing engagement, they did fill out a website URL, but did not enter anything in their profile description. It would be useful to have such accounts show up as suspect too - is there any setting that will do this?
I realise there’s no SEO penalties as these profiles are not linked. However, I do end up sending a fair amount of emails to these accounts and as I expect them to have very low opening/click rates they could end up hurting my email IP reputation. I prefer to run a tight ship and clean these up
This doesn’t fully address the issue you described but…
If most of these spammy / suspect users share the same stats you mentioned here
then, have you considered changing the default for the clean_up_inactive_users_after_days setting?
It’s set to 2 years unless you change it, but I see no harm in reducing that by quite a bit.
It only deletes users with 0 posts and 0 reading time.
Setting it to something like 2-3 weeks is very reasonable. Anyone that signs up to a Discourse site and posts / reads nothing for 2-3 weeks is obviously not a real person that you want to have on your community.
Even if they are, 2-3 weeks is an acceptable amount of time for their account to be deleted because they probably lost interest and so it’s okay for them to have to go through the signup process one more time if they are really interested.
If the commit is in tests-passed branch then it can be deployed on instances that are tracking tests-passed branch. (by default all instances track tests-passed branch.)
I see the commit is already present in tests-passed. So yes, the fix should be available for you to deploy.
