April 23, 2014

Boston Strong

The IBI Team

Topic:   Data

Massachusetts, Maine, Wisconsin (and maybe some in Florida according to Wikipedia) celebrated Patriots’ Day this week, a civic holiday celebrating the Battles of Lexington and Concord.  Personally, when I think Patriots’ Day, I think about Boston and sports, and unfortunately, after last year’s marathon, there will always be another memory.

I’ve been lucky enough to be in Boston for Patriots’ Day on more than one occasion and it’s definitely an experience I will not forget. I highly recommend it if you ever get the chance to go. You really get an understanding of the passion that the people have for the city and their sports. 

To dig a little deeper into the Boston spirit, I’ve used the technology at my disposal here at Information Builders to pull the conversations from the Boston Red Sox, Boston Bruins, and Boston Marathon Facebook pages, analyze them and visualize the result. Here’s a summary of what I found.

In terms of page “Likes”:

  • Boston Marathon has 187,373
  • Boston Red Sox has 4,535,478
  • Boston Bruins has 1,785,711

So, now you have a sense of the perspective when you look at the results for “share of voice” analysis on Patriots Day below.

In terms of “Share of Voice” (percentage of conversation):

  • The Boston Marathon had 18% of the conversation or 1,074 posts/comments.
  • The Sox had 38% of the conversation or 2,188 posts/comments.
  • The Bruins had 44% of the conversation or 2,550 posts/comments

One of the analyses we always like to apply to social data is Sentiment scoring, which gives you a sense of how positive or negative the conversation was on each of the Facebook pages.  The clear winner of the Sentiment battle on Patriots day was the Marathon with 38.5 percent of the posts and comments scoring positively and only 2.3% of the posts/comments scoring negatively.  The Sox had the highest percentage of negative commentary at 9.2%, but if you think about it, that is surprisingly low on a day in which they lost (to the Orioles 7-6).

Another type of analysis we apply to social data (or any textual data for that matter) is Word Frequency Analysis.  This counts the amount of times a word is used in a set of text, in this case posts/comments on Facebook pages.

For purposes of this exercise we filtered out the word “Boston,” which as you would imagine was the most frequent word on each of the pages. After that, here are the most frequent topics on each of the pages:

  • For the Marathon:  “Marathon”, “Congratulations”, and  “Runners”
  • For the Sox: “Strong”, “Pedroia”, and “Awesome”
  • For the Bruins: “Chara”, “Detroit” and “Wings”

One of the words that you’ll find in tag clouds for each of the pages is “Strong”, which maps to the  “Boston Strong” rallying cry in the city after last year’s terrible tragedy. Having been in the city on Patriots’ Day’s in the past, there was no doubt in my mind that the city would respond this year stronger than ever before.

It is always an interesting exercise to apply these analysis techniques to social data that maps to pop culture or the things that are going on around me in my life (see posts for Elf on the Shelf, Candy Crush, the Oscars, and March Madness). From a business perspective, our customers are getting significant value when applying these techniques (word frequency and sentiment) to the likes of call center data, patient surveys, citizen feedback, or quality notes.

If you want to learn more about this technique and low to leverage it, I am hosting an upcoming webinar on the topic of “Feedback Analytics” which will go more in depth into the use cases. You can sign up for it here: