I still do not know if it was due to my 24 tweet in a day or my follow/unfollow bot...
_ The follow/unfollow bot is clearly a risk, if the one you use has options like quantity | hours to run task | delay between 2 un/follow, use them to avoid flags.
In the first weeks of my account, it's been blocked a few times because I was auto un/following (remote app, using API). The default interval was 5 min. I tried 6, 7 & 8 : same result. But when I changed to 10, no more problems.
_ 24 tweets /d are not so much. But if 20 of them contains urls that looks bad for Twitter's eyes...
(If you short them, check some of them from time to time. Bitly (or others) can flag some urls as dangerous and block them without telling you.)
If your tweets with links are not back in analytics, they are probably the reason.
If they are, maybe your bot needs education.
(...) I think you might be onto something with this one, it's not the domain itself and you need to remove the tracking part only.
Yes, I can use the domain, it's a good thing and the more important (less things to modify).
The loss of tracking possibility is less important, but still a major problem to me.
When the landing page will be ready, maybe I'll try to use other names more currently used (like
page=,
post=,...) and forward them to
track= in the url to CB.
Or, I just thought, simply use an url form like
domain.tld/maybeARealSlugHere/trackingInfosHere/ .
I like to use the slug part because when twitter replace the link, the url is truncated (at ~25/27 chars) and the tracking part is not seeable in the tweet (replaced by "..."). I like to think it's more "click friendly".
