Those of you with a Twitter advertising account may have noticed that their analytics system has had an update this week, and I’ve already seen several posts talking about the impact this will have on social marketing.
I thought it was worth checking out the changes and seeing if they will in fact lead to a social data revolution…
The short answer is.. nope. This is essentially an aesthetic update, with little impact on your strategy. However, it does have some interesting figures included that will make optimisation faster and easier for Twitter managers.
Before we dive in, if you don’t have access to Twitter Analytics, I’ve previously written a simple walk-through guide. It requires you to register for an advertisement account, but don’t worry, it won’t cost you anything.
Now, on with the updates, which visually seem to have taken a leaf from Facebook Insights:
(Incidentally, I’ve redacted some of our info here, because hey, I’m not just handing benchmark data out to competitors. Also, I’ve been on holiday for a week so things have dropped off a bit over the last few days, I don’t wanna look bad y’know!)
A focus on reach
At the top of our analytics page we’re presented with a nice, clean bar graph showing engagement over the past 28 days.
Currently there’s no option to extend that date range so you’ll need to bookmark it regularly.
It’s interesting that Twitter is highlighting ‘impressions’ here, which ties in well with its push to engage TV audiences.
In broadcast, reach counts (in many other cases though, it can be what we technically term a ‘bullshit metric’). On the plus side, it’s a lot easier to see than the old dashboard, so you can quickly identify peak days, helping you learn when your audience is online.
Does anyone really care?
Underneath this you’ll see a list of your recent tweets, with information on impressions and engagement which is again useful for optimisation.
At the side of the page we’re presented with smaller charts detailing Engagement, Link Clicks, ReTweets, Favourites and Replies:
Again, these are all useful to have an idea of how your audience interacts with your content, and identifying strengths and weaknesses.
So far, this is all vaguely useful, if not life changing stuff. The real meat of the update happens when you export this data:
A CLV will give you info on all sorts of things, but one I found interesting was the separation of clicks. Rather than just providing a single number, you are now shown whether a user clicked on an @username, a URL or a hashtag within the tweet, as well as whether a user expanded or clicked any embedded media from a tweet.
For Econsultancy this is useful, as we can now see if people are really reading those charts we post, or if they prefer big colourful pictures of brands, among other things.
Image optimisation is increasingly important as Twitter users become more comfortable with posting a variety of media. The stream is becoming cluttered so this can help you stand out.
Improved multichannel measurement
Another huge improvement is the attempt to measure beyond the immediate Twitter ecosystem. New metrics now display occasions where a tweet was emailed. This is a going to be a tricky one to track perfectly, but it may give further insight into just how much these channels bleed into each other.
There’s also a ‘dial phone’ link. This isn’t something we use at Econsultancy, but should you be tweeting out a phone number for any reason, this will let you know whenever someone clicks and calls using a smartphone.
Finally there are also metrics for Twitter cards. While these have massive potential, they are currently under-utilised, partially due to some rather odd formatting requirements (Our own experiments showed that they tended to have a low reach, but did encourage high response rates).
All in all this isn’t really groundbreaking, but it certainly gives you more useful insight in a more presentable form, and is particularly useful for pushing up engagements and identifying value.
If your users love hashtags, you can start to work out which ones are driving business for you, rather than just cramming more in to every tweet and hoping against hope.
As usual I’ll get myself off the hook by ending on a question: What would you like to see from twitter analytics? Which metrics are most useful? It’s always fascinating to hear from marketers in different sectors so please do let me know in the comments.