I have a bot that replies user with a link of media content contained in the parent tweet when they mention me as a reply. Everything worked fine for the first 2 days until it just stopped. Currently, users cant see the response from the bot in response to them, they can see there is a response but expanding tweet reveals nothing. It’s more interesting that I can see both the user mention and bot response from the bot’s account. Does anyone know what the likely cause is and how I can resolve it?
It looks like Sniper (or another spam detection system) is marking your bot’s tweets as spam. Do you know if people are blocking your bot or reporting it for spam?
As for solving this, it’s not easy considering your use case. I think you need to avoid making your bot’s tweets seem like spam. Are you replying only to the user that mention you or other users too? Is your content duplicate or similar?
Also, you might have to start with a fresh account that’s not marked for spam.
Thanks for the response. As far as i know, i don’t think people are blocking my Bot as its a utility Bot and it only works if you mention it in a tweet. I only reply to the user that mentioned the Bot, using the Id_str property returned from the tweet that the Bot was Mentioned
I send out generic tweets to everyone in response to their request, the only difference would be in the link to the particular media they are requesting
A typical response from Bot would be “Hi @twitterhandle, click on [media link to download] to download. You can always visit [external page link] to download your recent videos”
Do you think I am doing anything wrong with this?
I don’t think you are doing anything wrong per se, it appears that you are following the automation rules.
But like I said it is tricky since the tweets might appear like spam. Can you try changing the wording in your tweets and see if that works?
Not sure if whitelisting your bot is a possible option, perhaps one of the twitter staff here will give you better feedback than me.
So i followed your suggestions, changed the wording in the Bot response tweet also moved the @ of the user making the request to the beginning of the Bot response and everything has been working perfectly well for the past 1hr. I must say twitter API acts weird.
Thanks for pointing me in the right directions
You are most welcome!
The issue is not with the API but rather the spam detection algorithms, you can probably reproduce the same behavior if you tweet the same message from the web client
I am glad it all worked out in the end.