Shadow Ban Questions


#1

Recently a bot account I have got shadow banned and I was surprised how little info online I could find about that. Is it permanent or temporary? I’m currently testing it by doing a search and seeing if my bot shows up the results.

The bot is called @Aloebud and it:

  1. lets people tend a garden by sending emoji resources to it
  2. send digital flowers to people

the way the flower sending works is you DM the bot and then it tags the person you want to send the flower to and the person sending it. It’s kinda cute and has very high engagement.

but I guess it got too popular because we seem to have gotten shadow banned? We are not appearing in any search results and TBH I’m not sure what else shadow banning does…

anyone have any experiece with that? Or the best way to ask for a 2nd chance from twitter?

I did submit to https://support.twitter.com/forms/platform and got a responce (I think automated) that my app keys were re-enabled… but they were never fully disabled… so that didn’t seem to be the answer.

I do understand that shadow bans are to stop abuse. I don’t think letting people send flowers to each other counts but I guess I’m open to being wrong :slight_smile:


#2

You may not actually be shadow banned. You mention using the search api, the problem with doing that is that the general search api only returns up to 1% of the phirehose. If your query has too many results you will not get them all which may be what is happening.

Also, the way I understand shadow bans to work is that your content is only visible to you so that you don’t even notice you’re shadow banned without some work. That’s kind of the opposite of what you’re experiencing as you can’t even find your own content yourself, as far as I understand.

It is possible you trigger some automated flag like you’re worried about though, I just kind of doubt it from what you’ve described. In any case, you aren’t likely to get very many specific official answers from this because if you did indeed trigger something then Twitter staff would be revealing too much information to possible abusers by answering you if you did in fact trigger something.


#3

Looks like a cute app!

I’m not able to comment on specific details of Twitter’s anti spam algorithms, but in general I’d direct you to the automation rules which provide guidance on behaviours which are prohibited.

One thing I can see that might cause issues in what you’re describing is that your DM-to-send-flower-to-friend mechanic does not actually cause the recipient to opt-in to receive the @mention, which would therefore count as an unsolicited mention. If the recipient (or others) chooses to mark those Tweets as spam, or generally your app is sending a number of unsolicited replies with which the recipients never interact, then this would be a signal of unwanted behaviour.

We are unable to help with individual app questions here, the best way to raise queries is via the support form you’ve submitted.