A statistical profile of how Twitter users engage with Dynamic Ecology

We interrupt your regular scheduled important content for boring trivial content. Below is a quick statistical profile of how Twitter users engage with @dynamicecology. It’s mostly just me making notes to myself, but in the unlikely event you care too, have a look.

Here’s a monthly time series of our total number of Twitter followers, from Sept. 2016 up until Nov. 2018 (when I downloaded these data and then promptly forgot about them):

Slide2

Ongoing steady linear growth. Not quite sure if I should be surprised or not, given that our posting rate (and hence the rate at which @dynamicecology tweets) hasn’t changed or has dropped slightly over the years.

The next graph gives monthly time series on how those followers (and other Twitter users) engage with our tweets. All measures of engagement are expressed on a per-follower basis to correct for growth in our follower count over time:

Slide1

The take-home points:

  • I looked up these data because I’d had the anecdotal impression that our tweets were drawing less engagement than they used to. I was wrong. On a per-follower basis, there’s no long-term trend in any measure of Twitter engagement.
  • Only a small number of people engage in any way with any given tweet of ours. Even over the course of an entire month (during which time we publish 15-20 posts), we never get more than about 0.75 link clicks/follower. That is, the average follower of @dynamicecology clicks less than one link in our tweets per month. I strongly suspect that’s because a very small minority of our Twitter followers click through to many of our posts, while a large majority basically never click through.
  • Different measures of engagement are all positively correlated with one another. I haven’t shown it, but the same is true on a per-tweet basis. Tweets with links that many people click also tend to be tweets that get liked, retweeted, and replied to. Unsurprisingly.
  • Link clicks are by far the most common form of engagement with our tweets, followed in descending order by likes, retweets, and replies. None of which is surprising, and is not because our Twitter account is just a robot to which there’s no point in replying. Rather, reply tweets are rare because only a small minority of Twitter users use Twitter for conversations. For instance, the vast majority of tweets about scientific papers are broadcasts rather than conversation/discussion: people, journals, and bots tweeting or retweeting notices of papers, with no substantive commentary.
  • There’s no obvious signal in these data of changes in how we use our Twitter account. Our Twitter account has always mainly been a robot that just announces our new posts with links to them. I used to use it occasionally to tweet myself, mostly to joke around with @duffy_ma but sometimes to discuss our posts with others. I stopped tweeting because Twitter discussions don’t work for me (YMMV, of course). But you can’t see any change in these data since I stopped tweeting. Presumably because I never tweeted much in the first place and didn’t get much engagement when I did. So hardly anyone noticed when I stopped.

7 thoughts on “A statistical profile of how Twitter users engage with Dynamic Ecology

  1. Congratulations to all you for an incredibly successful, thoughtful, regular, and interesting blog. I don’t engage as often as I would like, but I would have to say that about pretty much everything that interests me!

  2. Honest question for our many Twitter followers, from someone who doesn’t use Twitter: why do you follow us? I’m especially curious about this if you, like most of our followers, rarely or never click links in our tweets.

    Is it as simple as, you once saw a tweet of ours you liked, so you decided to follow us, and then just never bothered to unfollow us when it turned out that most of our tweets weren’t interesting to you?

    • Log into your Twitter account in your web browser. In the upper right, click on the icon with your avatar image. In the resulting dropdown menu, click “Analytics”.

      • Right but how did download specific data like number of followers, likes, engagements, etc for such a long period of time. I can only get those data month by month. Did you have to download them all separately?

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.