Since it’s been too long since my last post (and since I’ve accidentally arrived at City 2 hours early for my class), I thought it’s about time for a quick one about my first go at looking critically into my online stats.
In the past, I’ve always judged whether my posts were successful or not by looking at the fairly shallow metrics of retweets, replies and favourites. But the analytics rabbit hole goes much deeper than that.
I was introduced to Twitter’s analytics service in one of my social media classes about a month ago, and it’s been tracking my stats since.
Unfortunately, it only starts measuring once you’ve signed into it for the first time. So if you’re a newbie and want to see the performance of your last few months’ worth of tweeting, then you’ll be disappointed.
Other Twitter analytics tools like Twitter Counter can measure some stats from before your first sign-in, but you have to pay for most of the site’s premium features.
Still, if you’ve got some money burning a hole in your pocket and want to keep track of unfollowers, interaction rates with other users, and a bunch of other cool stuff, it’d be well worth it.
Mine’s been ticking over for 28 days at the time of writing, so I’ve got a fairly good time range to be looking at.
Anyway, let’s cut to the chase.
Here’s the first thing you see when you open analytics – a lovely little bar chart of your impressions, or the number of times other people saw your tweet.
This post is very much a baseline record for me to look back on – I don’t really know whether 11.8K impressions over a month is a high or embarrassingly low number, but at least it’ll be interesting to look back on this in a few months to see how I’ve progressed.
In the second tab, there’s some analysis of your followers, with more lovely graphs.
This shows my increase in followers over the past year. That 100-follower spike at the end is when I started my course at City – clearly these social media classes are paying off.
At any rate, I got an average of 665 impressions a day over this period, and with my 421 followers, I’d say that’s an OK number.
These charts are fine if you’re curious about your overall performance on Twitter, but not so much if you want some really in-depth analysis.
Twitter Analytics works fine for what I use it for, but if you really want to do some number crunching, then you should probably start paying for another service.
Having said that, there is a handy feature that allows you to download all these metrics as a .csv, which is good if you want to get into some hot data spreadsheet action.
I’d like to crack into that once I know a little more about that side of things.
At this stage, Twitter Analytics’ most useful feature is the figures available in the ‘Tweets’ tab
Here’s one of my recent tweets (more of the usual indy stuff, I’m afraid), I thought it did fairly well at the time – a couple of retweets, some replies.
But to get a deeper idea of how it performed, you just have to click on the tweet and you get another lovely set of stats.
Also available behind this pop-up window is the engagement rate, which is the number of engagements (clicks, retweets, replies and things like that) divided by the number of impressions. This gives a handy figure that can illustrate your tweet’s popularity.
This one did fairly well, getting an engagement rate of 20.8%, way above my average of 4.1%.
Naturally, the tweet was all about the pictures, and the whole Alex Salmond effigy thing was causing a bit of a stir on Twitter at the time anyway.
I tried to add a little bit of commentary on top, which may explain why it did better than average.
Plenty of people had seen the tweet already from other users who tweet about Scottish stuff, so maybe my decision to rip off Limmy and stick #bettertogether on it made it a stand out a little bit.
Anyway, I’m trying to explain a joke here, which is never funny, so I’ll move on.
What makes some of my last months’ tweets more popular than others?
The success of the Salmond one is obvious – it’s topical, funny, and it’s got a hashtag in there.
But what factors made this stupid tweet:
So much more popular than this slightly less stupid one?
It’s tricky, and trying to figure it out is proving fairly difficult.
The tweet about my dad’s cool trainers got an engagement rate of 18.1%, and was seen 210 times.
The one about the crazy basketball shot was seen 216 times, just as much as the other one – but only had an engagement rate of 2.3% – basically, this tweet only got 3 clicks, compared to around 30 clicks for the other one.
On the face of it, the video of the basketballer should attract more views. It’s got a fairly intriguing title, and a bit of commentary from me. It’s a fun video, check it out.
It’s great, right?
But somehow, a stupid tweet referencing a weird internet joke with a picture of my dad’s shoes attached was so much more popular.
I’m guessing this basically boils down to how we use Twitter. The vast majority of time I spend on Twitter is on my phone – I only really use it on my PC when I’m checking Tweetdeck in the morning.
The mobile effect.
Certainly if I was browsing Twitter on the bus or in class, I’d be far more likely to engage with a quick, throwaway gag tweet like that, than I would with a tweet that makes me click a link, go to a website, wait for a video to load and then watch the video for 2 minutes on mute.
It’s not entirely down to the actual content, but more down to how people use Twitter.
That becomes more obvious when I look at my tweets that link to articles. The link in this tweet got 10 clicks – not a huge number, but 3 times more than the basketballer one.
I’m guessing the reason is that people using Twitter on the bus or on the toilet (come on, we all do it) are more prepared to have a quick look at a tumblr blog than they are to watch a video, and that makes total sense – I’d do the exact same thing.
So do I need to change anything?
Maybe I should start thinking about timing more. I think if I tweeted links to videos like the basketballer one in the evening, when people are vegging out with their laptop, maybe more people would engage with it.
Similarly, if I focused on posting throwaway pictures rather than links to articles I liked, I’d probably get more engagements.
But at the same time, I don’t want to be a slave to analytics, constantly trying to push that bar chart at the top higher and higher.
This quick exercise in metrics has made me think about what I tweet and when, and whether I should be wasting my time posting lengthy content in the middle of the day when no-one’s going to bother looking at it.
Like I said, it’ll be useful to do one of these again in a couple of months, not only to see how those bar charts have changed, but also how I’ve altered the content I post.
I’m still fairly new to this social media game, and I’m trying to learn. If you’ve got any thoughts related to what I’ve posted here, then do let me know in the comments and we can have a chat.