To prove the connection between confidence and success, we created a social platform called “Most Confident Fans” to find the NFL’s most confident fans and teams.

But how do you accurately assess the confidence of a fan or an NFL team?

We combined the way fans voice their unfiltered passion and support of their team through their social posts on Facebook and Twitter with the cognitive abilities of IBM Watson.

We developed a proprietary algorithm that worked with Watson’s Personality Insight API, and had Watson analyze the words and tonality fans and players used to support their favorite NFL team. We were able to determine a fans' confidence level in their team and correlate that confidence to that teams predicted success each week.

Fans were motivated to improve their scores by a leaderboard that was updated in real-time…competing for the ultimate prize - a trip to the Super Bowl.

We developed a proprietary IBM Watson algorithm for what we defined as “confidence.” Next, we aggregated all the tweets from players on each NFL team, and ran them through the algorithm to determine a player confidence score. In addition, we gauged fan confidence in each team by running fan social posts through the same algorithm. Once combined, we had a confidence score for each team heading into every matchup that was used to predict the outcome.

We launched the program by organically reaching out to fans on Twitter that had a history of posting confidently about the NFL. We invited them to put their confidence to the test on our platform. After about a week, we then launched some light paid social support more broadly. The platform was live for a total of 5 weeks.

This initiative was promoted organically, attracting thousands of participants, The average participant visited 75% of the site. The average user spent well over a minute on each of the pages that we created, leading to a record breaking 20 minutes of engagement with the site.

The promotion drove over 212,000 organic impressions across Facebook and Twitter, and 3,000 organic social engagements.

The #MostConfidentFans hashtag was used 2,600+ times from Sept. 1, 2015 - Jan. 31, 2016.

#MostConfidentFans was the second-most used hashtag in conjunction with #SuperBowl50, even beating out mentions of halftime performers including Beyonce and Coldplay.

As the first advertiser to use IBM Watson in a marketing context, We actually predicted the winner of Super Bowl 50 weeks before the game, based on the data, they had the most confident fans. More importantly, investments with TD Ameritrade increased $800 million dollars during the initiative.