Two weeks ago I started the second semester of the Masters in Data Science program and as part of it I am taking a course in Social Media Analytics. The first lab assignment for this course was on January 25 and the objective is to analyze Bell Let’s Talk social media campaign. Using a proposed tool called Netlytic (a community-supported text and social networks analyzer that automatically summarizes and discovers social networks from online conversations on social media sites) created by the course’s professor Dr. Anatoliy Gruzd I downloaded a tiny slice of #BellLetsTalk hashtagged data and created this super simple Power BI dashboard.
I have been wanting to play with Power BI’s Publish to Web functionality for quite some time and thought this was a great chance to give it a cool use. The data was exported from Netlytic as three CSV files and then imported into Power BI desktop. With the desktop tool I created a couple of simple measures (Total number of tweets and posts, Average number of tweets and posts per minute and so on) and then some simple visualizations.
Conclusions: This campaign was on fire on Twitter (277.78 tweets/min) compared to Facebook (4.57 posts/min) and Instagram; when considering number of unique users Instagram comes first, Twitter in a very close second and Facebook in a third distant position. Also, as in this dataset every Twitter user has on average more followers than their Facebook counterparts, the audience reach on Twitter is 3x (200% more) when compared with Facebook. Notably, Instagram users were much more inclined to ‘Like’ posts from this campaign when compared to Facebook users: for every Facebook Like, there were 5 Instagram Likes.
A couple of minor side notes:
- as expected most of tweets and Facebook posts where in English and French
- most of Facebook post types where Status updates (the typical post in users’ Wall)
- majority of Instagram posts didn’t use any photo filter at all
This serves the purpose of demonstrating how much very simple it is to create Social Media analytics today using free and publicly available online tools: all the information is out there, very good and free tools are there too, you just need to be willing to do the work.