1

Hi. My application reference is Case# 0227905135. My bot idea was completely harmless. I’m sure a company as big as twitter could effortlessly turn the rejection reasons into a cheatsheet and turn headers into a rejection codes.Instead of sending “rejected” mail, with this way we can get more detailed information about rejections. While helping us to understand the values ​​of Twitter, we will have an idea about what we should pay attention to in our future attemt.

Thanks.

2

I can’t speak for Twitter on this, but I would bet that an awful lot of “harmless” bot ideas are not, in fact, harmless from a social media perspective. What was the use case?

3

Hi, My bot was a reminder. For example, when a company or publisher shares a tweet in proper template such as “gmt 2 3:00 am” on twich or someting,I was aiming to DM to caller(who invite bots to comment), 1 hour before the event start, according to his/her local time.

4

Hmm. Such a use will almost certainly get suspended, imo.

DMs are meant for synchronous conversation - for example, the user sends a message and your bot sends an immediate response in reply to that message. Your use case is asynchronous - i.e. the user sends a message, and your bot sends one or more messages at a future time.

There is more in the Automation Rules, but in general:

  • When sending automated replies (and DMs), you send only one reply or DM per message the user sends you
  • You must have the user’s consent, signified by the user sending you a request, which means getting consent each time before you send a DM, not once only

Your use case breaks this because it sends multiple DMs to a user who only requested once. The reason this restriction exists is because systems which say “we will message you forever until you opt out” are almost always spam systems that make for a bad user experience, and because user consent is not an infinite thing and should be requested immediately before each automated interaction.

Trying to make a bot that alerts users to events is not possible within the rules, so I suspect that’s why your application was rejected.

1 Like
5

Oh thanks for that information. Now i know whats wrong with my bot. You have any suggestion for bypass this problem ? I know my bot is not so useful but i really want to discover tweeter dev tools :smiley: If the bot tweets to comment section 1 hour before the event (instead of DMing the users), can I solve the problem? (like comments to anime bot )
Sorry for taking your time.

6

It’s no problem ^^

I don’t think it can work - the only way you can really do this is to have people following your bot, and then the bot tweets on its own rather than commenting to a user, for instance. There isn’t really a way to make it ‘notify’ the users outside of them seeing your bot’s tweets in their timeline that won’t break the rules.

Probably the best thing you can make is a tool that e.g. does a quote retweet (or just directly retweets) the event tweet, like “this is starting in 1 hour” followed by the main tweet or such like. Users who follow your bot would then see this in their timeline.

7

If you had a separate website, you could ask people to add their @ handles to specific reminders. that allows it to be opt-in and changeable at any time. Or, another way to solve this would be to have your bot tweet a header post for a specific event and ask people to comment below with a special keyword or hashtag (example: #RemindMe ) to be included in a reminder DM or post. As long as you allow users to opt-in first before sending anything and make it a one-time reminder not an ongoing thing, you should be fine.

8

You can’t do it this way (gaining opt-in on a separate website). Twitter has no way to know the user opted in on said external website, so it will look like an unsolicited mention.

Consent for DMs or replies is generally in the form of “the user DMs or tweets @ you on Twitter, and then you send one DM or tweet back to them”.

All of this being said, the rules are partly aimed at preventing “spammy or aggressive” mentions; sticking to one mention per user message might be fine, though I’m not sure if Twitter will necessarily like that if it’s done for many users.

Additionally, unlike other use cases, your bot will likely be sending out huge numbers of mentions at one time, which at least in my opinion is likely to raise alarm bells on Twitter’s side.

2 Likes
9

Yep, Ive battled that particular problem (separate website) and have a compromise of sorts in place that seems to be working. I’ve also slowly built up the threshold for these over many month so the twitter traffic cops dont get overly alarmed with the number of @'s it sends out. Every post my bot sends with an @ mention also includes a link to my sites TagMe page - just like the Unsubscribe link in emails - which allows them to opt-in or opt-out at any time. Finally, it never @'s someone who isn’t following. This walks the line of being spammy but does so with clear limits in place. What I’m finding (and what Twitter is hopefully finding) is that this service is adding real value to twitter as a platform for artists to thrive on. The bot has gone from 0 to 15k followers in 8 months or so.

10

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.