Comparative analysis of the conversation on NHL Team Facebook pages
With the puck set to drop on the 2014/2015 NHL season, I thought it would be a good time to do some social media analysis of the conversation on each of the 30 teams Facebook pages leading into the season.
Those of you outside New York may be unaware of this unofficial and unscientific cluster analysis. Ask any New York sports fan which teams they root for an more often than not you’ll get one of these two clusters:
· Cluster 1 – Football: NY Giants, Baseball: NY Yankees, Hockey: NY Rangers
· Cluster 2 – Football: NY Jets, Baseball: NY Mets, Hockey: NY Islanders
Unfortunately, I fall into cluster #2 (which could somewhat explain my haircut in the above picture). And as an Islander fan, preseason games are very often the only meaningful games that I get to watch. That's why I chose to do this analysis before the actual season starts.
For this analysis I used WebFOCUS’s ability to directly access the Facebook API and then applied a number of the platforms analytic capabilities, including the ability to quantify unstructured posts and comments with word frequency and sentiment analysis and then visualize the results.
We pulled posts and comments from 9/23/2014 to 10/07/2014, so essentially 2 weeks of posts and comments.
Across the 30 team pages there were 93,859 posts and comments during that time period
That conversation was made up of 64,486 unique individuals who made at least one post.
Those 93,859 posts generated a total of 2,570,788 likes
The most liked post was by the Detroit Red Wings on their own page and it generated 31,039 at the time of this analysis – “On this day 23 years ago, Nick Lidstrom and Vladimir Konstantinov made their NHL debuts against the Blackhawks. They combined for six Stanley Cups and gave all of us countless memories. #RedWings”
During our analysis, the most active day was 9/24 with 8,894 posts and the least active day was 10/1 with 167 posts.
In terms of Share of Voice, the top three teams were:
Red Wings at 8.39% of the conversation (7876 posts)
Toronto Maple Leafs at 6.6% of the conversation (6192 posts)
LA Kings at 6.24% of the conversation (5859 posts)
The bottom three teams were:
Florida Panthers at .93% of the conversation (876 posts)
Arizona Coyotes at .95% of the conversation (891 posts)
San Jose Sharks at 1.31% of the conversation (1253 posts)
That is a partial measure of engagement but I also like to look at the data a little differently and take the total number of people who like a team’s page and divide the number of posts by that to get a sense of scale. Some teams just have a far greater following. So this score is # of posts/# of people following the team page. Using this metric the LA Kings (.0067) come out on top in terms of creating an engaging conversation with their fans, followed closely by the Florida Panthers (.0065).
That is a lot of information, but it is all quantitative, in some social circles it would be classified as “engagement” level data – if, when, and how much activity took place. One of the advantages that social data provides is the volume of qualitative data it presents in the conversation. To harness this data you have to apply techniques like word frequency and sentiment scoring, which we have here. Let’s take a look at some of the qualitative results.
Across the 93,859 posts on all 30 teams pages the most popular words were somewhat expected:
Awesome – that makes you feel good if you’re in the NHL’s marketing department.
From a sentiment perspective, 26.97% or 25,314 posts were scored positively by the engine versus just 8.25% or 7,742 being scored negatively. Here are some team sentiment scores:
The Edmonton Oilers were the big winners in the sentiment scoring category with a total 54.18% of their posts and comments scoring with the engine. 46.89% of the commentary on their page scored positively versus only 7.29% negatively.
The Boston Bruins were the losers in the sentiment department with a whopping 19.32% of the commentary scoring negatively. That is a bit on the high side and will require some deeper analysis.
As you might expect from the current Stanley Cup Champions, the LA Kings had the lowest percentage of negativity on their page at only 3.93%
These are just some of the metrics you can gleam from social media data to help you get a better understanding of what your customer or fan base is saying about your brand or team or how they are responding to your marketing efforts. How you apply these metrics and integrate them into your business is really going to determine their value. Take a look at our recent blog post that outlines Four Social Media Analytic Use Cases that have Impact if you are interested in how you can take these metrics to the next level.
Best of luck to all your teams this season. This is the Islander’s year, I can feel it -- they are going to close the Nassau Coliseum with a playoff run.
Information Bloggers is a conversation between people at the nexus of business analytics, mobile apps, cloud computing, social media, and big data. Together we are developing information management strategies that change behavior and transform business. Many of us work for Information Builders and, as such, are passionate about customer satisfaction and dedicated to excellence. Our views are our own. Welcome to the discussion!