So I’ve been doing a bit of an experiment on Twitter recently. People talk readily about how using media-rich tweets improves engagement, there are blog posts and infographics devoted to the topic, and as someone who dwells in social media, I’m all for upping engagement. However, I wouldn’t be a scientist if I didn’t test the theory myself before adopting the practice. After all, if it doesn’t work for me, what would be the point in devoting time to doing it? Therefore, in the middle of April, I decided to up my posting of media-rich tweets and see what, if anything, would happen.
To do this I used Hootsuite and my marine science-based Twitter account. I mentioned a while back that I’d only just realised you could attach images through Hootsuite, and it seemed the appropriate time to make use of the feature. Usually, I would use Buffer for media tweets, but as I’m limited to just 10 posts with the free account, I opted to use Hootsuite. I’m still limited to 10 tweets a day (if I just autoschedule posts), but not in total, so any extra tweets rolled simply over to the next day. That meant I had Buffer free to retweet other people’s media tweets. So, overall, my number of tweets per day didn’t increase much, the tweet types just switched platforms.
As for images, well, I’d usually included an image with a tweet if the article in question had a nice image of a marine mammal, but for my test, any news article I came across that had a nice looking image included, or was one that I could attach a CC00 image to, I made into a media tweet. So queue pictures of fish, coral reefs, and journal diagrams. I didn’t use any icky pictures, such as stranded dead dolphins. Those articles I just scheduled as is, and if a thumbnail popped up in my feed with the image, then so be it. Then I just tweeted away….
Come the end of April, I downloaded the analytics from Twitter, picking the date range so the 12th was in the middle, and had a butchers at the figures. And what did they tell me?
Well, first off, Twitter gives you a load of graphs on its Analytics page, but as someone who’s a bit knew to the whole analytics game, I have no idea if I’m reading them correctly. The first is the big impressions graph….
Here, like the other graphs, I’ve marked the 12th in red. Now I’m assuming ‘impressions’ just means ‘people who’ve seen the tweet’. So the more something’s been retweeted, the more people who will have seen it. From a glance, it does seem that after the 12th my impression level went up, but this could also be due to the fact that I had a couple of tweets retweeted by people who had big followings themselves, thereby bumping up my figures. But then there’s the question: if they hadn’t been media tweets, would they have still been retweeted? Either way, as they were media tweets, I’ll take it.
Twitter also provides a number of smaller graphs….
For the most part, they don’t really help prove the hypothesis that media-rich equals more engagement, with the possible exception of the ‘Likes’ and ‘Retweets’ graphs, which do seem to show a slight increase in interactions with my posts. ‘Link clicks’ and ‘Replies’ don’t really show an increase in interactions. ‘Engagement rate’ is just all over the shop, and I can’t figure out if it shows an improvement or not.
However, Twitter analytics also allows you to download the data so you can export it into Excel (or your spreadsheet program of choice). The data was a bit tiresome to trawl through, but for the sake of science, I persevered. I did some averaging out, and this is what I got (see table below). Again, I’m making an assumption as to what ‘Impressions’ and ‘Engagements’ mean, and I have no idea how ‘Engagement Rate’ is calculated, but that being said, there does appear to be a noticeable increase across the board after I upped my media tweets. Now, whether that increase is statistically significant is something I can’t be bothered to work out.
|Impressions||How many people see tweet?|
|Mean 25th March-11th April||112.230320699708|
|Mean 12th April-30th April||146.84487534626|
|Engagements||Likes, retweets, link clicks per tweet?|
|Mean 25th March-11th April||1.57725947521866|
|Mean 12th April-30th April||2.29362880886427|
|Mean 25th March-11th April||0.011296937684149|
|Mean 12th April-30th April||0.013246888306627|
I even did a graph for daily Engagement Rate….
Now, I’m not sure why my engagement rate graph came out different to twitter’s, but it does help show that there was an increase after I upped my media tweets.
So, what can we surmise from all this? My little experiment may have lasted only 18 days, but even so, it does seem the general consensus was correct; more media-rich tweets leads to increased engagement.
Therefore, I shall continue doing what I’m doing. And if you were in two minds as to whether to do the same to improve your own Twitter account, hopefully my results have helped you make a decision.