Twitter Analytics with Qlik Sense

Since a few months Twitter allows their users to do more on analytics from their own tweets. Not only do they provide a download to your tweets and detailed statistics that come with that, they also provide insight into your twitter performance. Very cool I think. This is a nice step forward in the landscape of social media and social analytics.

This got me thinking about doing my own point of view on top of that data, I mean the raw data that everyone can download on http://analytics.twitter.com. So this led me to the following table of content for my own and personal Qlik Sense dashboard.

1.   Dashboard 

• The dashboard consists of several KPIs that tell me something about the interest of my audience and their engagement with my interests. KPIs on this screen are:
  • Total tweets
  • Total impressions
  • Total engagements
  • Total retweets
• Next to those KPIs the dashboard in the app helps me to get a visual hint on what my tweets look like on a high aggregation level:
  • My tweets over time give insight in my activity on twitter
  • The hashtags I used in my tweets are being visualized using the pareto (80/20 rule) principle, this will give me a clue on who I try to target and what my interests are
  • I also use a wordcloud to see what words I use in my tweets for all tweets I have been posting

2.   Interactions 

• On this interface I try to get insights into my interactions with other twitter enthusiast, peers or people who I would like to have a conversation with every now and then. On this screen you’ll find visuals that help me to understand:
  • With which people I interact the most and this is layed out for me in a nice barchart
  • Next to that you’ll find insights in who is retweeting me and how much, are there any fans already? Would be nice to eventially come up with a definition for the word ‘fan’. Right?
  • Then I also created an additional chart to support the intensity for interactions within my network. I like to call this my hairball with interactions. The bigger the knots in there, the more interactions I had with these people (or nodes as our techies would say 😉
  • Do to the interactivity this application provides I can do some nice ‘point and click’ analysis. The beauty of this is that all objects will immediately adhere to my selections. For this reason I also already give some more detailed information in tabular format

3.   Performance scan 

• The UI for this helps me to dive into the performance, relations and proportions of my posted tweets. This will help me to find out if certain hashtags or times of day are better for engagement or not. Next to that ‘point and click’ analysis will also get met to detailed or tabular views of my tweet stats. These insights will address ‘actionable information’ and what the details are for my selections. These details can be found on the last sheet in this app, called ‘Tweet update overview’, the more granular level of details for my tweets.

“Ideally when you want a ‘big’ overview of everything you would have a look from 30,000 feet high. But then you do not see the details too good. This is why you need to have multiple layers and altitudes in your views and visuals on data.”

Would you like to have these same insights?

You can now download this dashboard in the QlikShow App Store.

Dashboard with all your important metrics and visuals on 30,000 feet

Interface with who you interact and how much

And last but not least, how are your tweets performing?

Please leave some comments when you have questions, just want to support this initiative or have an opinion about the app.

Stay tuned if not 😉 …