This is something I think about WAY too much.

So I always think of a good tweet but then second guess the way I should phrase it, or the caption I should use.

Here’s an example from last week:

So I thought this was gas, and admittedly it got a decent number of retweets and likes, but it could have done SO much better with a decent caption. I toyed with one for ages, like “When it’s time to Rise & Shine” or something punny like that, but nothing I thought was universally relatable so I didn’t want to risk not getting retweets based on the caption.

A/B testing would be gorg in this instance, I could have done one tweet with the caption, one tweet without, and then I’d know which one would do better.

It would also be good from a timing point of view, to see which time would be best to post your tweet at to get maximum engagement. Although I guess at this stage most active Twitter users are aware of this anyway, and Twitter analytics does a pretty good job of showing you peak engagement times.

Of course there’s like a bazillion problems with this. I guess first of all, A/B testing would technically only vary the users who see my initial tweet, but then RT’s of both of the tweets could be seen by any user so there may be some cross over and the results might not be super accurate.

I’m not sure if this is a service available on Twitter for Business, but from a quick Google it seems it’s only something you can do through a third party like Bitly. Even doing that they basically just recommend sending out two variations of the same tweet to see which performs better, which idk sounds like the worst and most inaccurate way to collect data ever but ok.

I guess Jack has other things to think about this quarter anyway. 🙂