Putting Twitter Analytics to the Test
We blogged not too long ago about the news that Twitter was beginning to roll out an analytics dashboard for everyday users, offering helpful information on things like followers and interactions that had previously only been available to Twitter advertisers. As digital strategists for a number of different clients, we explained the value in leveraging performance metrics to guide content strategy on Twitter, citing the importance of continually evolving your content based on what you are seeing working and not working. Well, we decided to put our money where our mouth is and see what the analytics dashboard had to say about our own content for @LPP_PR, hoping to draw some conclusions about where our strengths lie and what improvements we could make.
We chose to look at what Twitter called out as our “Best” and “Good” posts of the past month, as that offered a large enough sample with enough of a variety of posts for us to make an assertion or two, though to more thoroughly compare posts we’d recommend doing formal A/B testing over a longer period of time. A/B testing on Twitter would involve sharing a particular piece of content several times, tweaking it slightly each time to see which presentation comparatively produces the most engagement. There is a great variety of ways to present any piece of content, and using A/B testing with Twitter’s analytics dashboard to compare each should allow you to see what’s working best.
Our analysis of @LPP_PR content produced three key traits for what Twitter considers our “Good” and “Best” tweets of the past few weeks:
- Educational: Approximately one third of our most successful posts had a strong educational element to them, often in the form of a series of tips or a guide, among other things. The takeaway here is that content will resonate when you give back to your audience, offering them value they will not find elsewhere.
- Timely: More than half of our best posts were tied to something people were already talking about at that moment, either around a particular event that we were involved in or tied to some sort of news or pop culture event that was trending on Twitter. Connecting yourself to larger trends outside the scope of what you do can set you up for more frequent and more substantial engagement.
- Personal: Finally, more than half of our posts in some way probed our audience for engagement, usually by posing a question. While not every question is going to spark a major debate, inviting your audience to share their opinion or at the very least contest yours is a step up from simple broadcasting.
With such a small sample size it’s difficult to definitively say what practices are going to take a content strategy to the next level and which habits could sink a brand’s social presence, but at the very least we’re going to continue playing around with Twitter’s analytics dashboard to see what we can learn that we haven’t already gathered from third-party analytics services.
If you’re one of the chosen ones who has access to these analytics already, tweet us your thoughts about your experience so far at @LPP_PR (we’re actually interested, that’s not a trick to get our numbers up).