Social Media Swishes and Fouls in the Final Four®

Share with your network!

In the run-up to the first week of the NCAA basketball tournament, Proofpoint Nexgate researchers analyzed social media activity on official team Facebook pages and described both the winning energy and the increased spam and phishing that correlate to teams’ social media activity. Now that the tournament has finished, and Duke and UConn are crowned the Men’s and Women’s champions respectively, we can look back over the “social media tournament” to identify some winning and losing trends that carry lessons for any organization concerned about managing social media security and brand risk.

After week 1 of the tournament, we assessed the “rowdiness” of fans, as measured by the overall volume of posts on a team’s authorized Facebook page, and demonstrated how spammers and scammers are drawn to social media brands with large numbers of engaged followers. The network effect of these numbers enabled a single inappropriate post to reach an audience of millions. Looking at the authorized Facebook pages of the Men’s Final Four teams, we saw this trend continue and accelerate as fans concentrated around the remaining teams.

Social media tournament

Kentucky Wildcats
In our last Nexgate post, we predicted that the Wildcats would win the championship based on the “rowdiness” measured during week one of March Madness. We found an interesting statistic in the data on the Wildcat’s accounts: while total number of comments decreased, the absolute number of comments containing profanity increased. In relative percentages, there was a twofold increase of profanity between weeks one and two. Furthermore, there was a fivefold increase in relative hate speech between weeks one and two.

For an event that caters to alumni and families with children, it is highly inappropriate content and undoubtedly content that the University of Kentucky would not want associated with their brand. The rapid increase underscores the ease with which a conversation can be hijacked and a reputable brand can be put at risk by uncontrolled social media streams.

Wisconsin Badgers
More than any other team in the Final Four, we found that political speech on the Badgers’ pages diluted the content and messages from their authorized Facebook Brand Page. For instance, there was a considerable amount of content posted on the subject of Wisconsin Governor Scott Walker, much of it unflattering and distracting from the positive brand image of these tournament finalists. 

Figure 1: Inappropriate political content on Wisconsin Badgers Facebook page

Michigan State Spartans
Proofpoint Nexgate analysis found that Michigan State was the only team in the Men’s Final Four® that increased in rowdiness between Weeks one and two by a factor of 100%. Interestingly, their social media pages also decreased in profanity by 300%.

We also found that spam across the Final Four® was still a significant problem, and one that did not appear to abate as the tournament went on. Here is our spammer from our previous analysis, now taking advantage of the Spartans’ popularity and winning streak:

Figure 2: Unsolicited content (spam) posts on Michigan State authorized Facebook page

With no controls in place, unsolicited content like this is free to run rampant on a brand’s authorized page.

Duke Blue Devils
While spam continued to hit every team’s social media page, Duke had little else of note. It may be no wonder, as Duke also had four times fewer comments as the next least-rowdiest team.

To keep score of the rowdiest teams in the Final Four, here were the current rankings during the week of Sweet Sixteen® play, which selected teams that would advance to the Elite Eight®:

  1. Kentucky Wildcats (12,514)
  2. Michigan State Spartans (7,276)
  3. Wisconsin Badgers (3,255)
  4. Duke Blue Devils (828)

Proofpoint Nexgate also analyzed activity during the final week of March Madness, from the period of March 31, 2015 to April 7, 2015. We found that controversial topics on the Kentucky WIldcats’ page, which includes political speech and religion, increased by 400%, and adult language and hate speech increased by 100%.

Conversely, we found no significant change in social media activity for Duke, which was surprising given their win.

Taken as a whole, these trends highlight the brand risks that face every organization with a social media presence.

A winning formula?

In our analysis of data from the 2014 tournament, we noticed an interesting phenomenon: every team that had a net increase in fan rowdiness during the first week of the tournament advanced to the Sweet Sixteen. In other words, all five teams with positive “fan momentum” heading into the Round of 32 won their match and advanced to the Sweet Sixteen, and one of them ended up winning the entire tournament. Could fan momentum be a predictor of tournament performance?

Analyzing 2015 tournament data, we found that again every team with positive fan momentum heading into the Round of 32 advanced to the Sweet Sixteen®. The chart below shows the relative percentage change, with the positive changes (those points above 0) labeled.

Michigan State, West Virginia, Notre Dame, and Gonzaga all showed a net increase in fan engagement during the first week of the tournament, and – as in 2014 – all advanced to the Sweet Sixteen. While it is too early to assert that this metric is a predictor of tournament performance, we can see limits on it, to the extent that it only seems to relate to the teams’ performance in the Round of 32 – unlike last year, the overall tournament winner was not among this group. In 2015, performing the same analysis revealed that only one team within the Final Four® increased in rowdiness during the Sweet Sixteen and Elite Eight® matchups, and that team was Michigan State, who lost in the tournament semi-finals. We look forward to retesting the potentially predictive nature of this metric in next year’s NCAA basketball tournament.

Social media channels offer organizations a variety of powerful tools for engaging with customers and supporters, but as the experiences of these teams demonstrate, their popularity – coupled with a frequent lack of controls – make them ripe targets for scams, phishing, and hijacking by other parties and actors, all to the potential detriment of the organization’s brand. Organizations cannot ignore social media, but they also must embrace tools and techniques that enable them to define and enforce controls around content on their social media channels – there is much to win, and even more to be lost.

Subscribe to the Proofpoint Blog