20 December 2011 | Team Tamar

A sneak preview of Twitter analytics, thanks to @ThatKevinSmith

If you’re an astute Twitter user with your finger on the proverbial pulse, you may know that film director, podcaster and general raconteur Kevin Smith recently became a Twitter shareholder (albeit not, presumably, to Prince Alwaleed bin Talal levels). His link to the company, as well as his circa-2 million followers presumably makes him part of the “Power User” base – and seems to have lead Smith to receiving an early preview of Twitter analytics.

Speaking on his “SMornings” internet radio show yesterday, Smith revealed that he’d been given access to analytics on his account in the past week, and gave a very useful preview of what insight the package gave him. Recent previews (including this one from Twitter themselves) highlight the power of the tool for traffic analysis, but have stayed reasonably quiet on the “people” parts. So Smith’s revelations in this area should prove very interesting…

You can listen to the entire show here, but if you haven’t got the 2+ hours free to listen, here’s a summary of what Smith revealed on the show:

  • Where in the world? The package allows you to see where in the world your followers are – not just by country, but by both state and city level in some cases. No need for me to tell you how useful this data is for marketers and brands alike – as Smith himself pointed out, this data will help him to plan where his future live shows should be.
  • Who else do they like? Data about who else your followers follow will prove very useful to brands keen to find out what demographics their followers sit in. Incidentally, if you’re interesting in finding out how many followers you and another account have in common, Twiangulate is a great tool
  • What are they interested in? This is the one that intrigued me most, simply because it’s not clear exactly where this data comes from. Labels such as “Food”, “Politics”, “Campaigning” and (bizarrely) “Coffee” remind me a lot of the topics that tools like Klout use to categorise users. However, one has to assume that Twitter’s topic algorithm is stronger than Klouts – otherwise my own (not entirely accurate) influence in topics like “Unicorns” and “Religion and Spirituality” might raise eyebrows with some of the brands I follow…!
  • Boy or Girl? Gender is one of the most basic analytics stats you’d hope for, although not one that you’d immediately expect with Twitter. Why, I hear you ask? Well, simple – you never actually TELL Twitter whether you’re male or female during the sign-up process… so presumably this is guess-work?

By the sounds of it there’s a lot more included in there, including the aforementioned data about tweet popularity, click-throughs and follower growth. But the part that captured my imagination most was the “topic” segmentation – something I’ve always been very interested in. Segmenting users based on analysis of their content is a very tricky business, as some Klout users know very well, so I’m really intrigued to see how well Twitter manage this.

Of course, it may end up that this data is never actually made available on a personal level – whilst you may soon be able to get an estimate of what your followers are like, Twitter may choose to not make available the data it has on YOU, as a matter of security – or to simply avoid people spotting holes in the process. I guess only time will tell – but I for one am quite excited to find out…

The most recent preview of analytics, thanks to Twitter

Team Tamar