by Roman Vaculin
Want to know how far Novac Djokovic ran in his second-round match against Jarkko Nieminen? IBM’s SlamTracker knows (1,647.9 meters). While SlamTracker keeps up with every bit of on-court action at Wimbledon, this year my team at IBM Research added analysis of what happens online, off the court – even what’s going to happen – as part of the All England Club’s digital experience.
Our analytics, which is part of the Wimbledon Social Command Center, now pulls in social media data and identifies in real-time the most engaging and influential voices around Wimbledon and tennis. And our predictive analytics, using keywords, hashtags, and other online sources, helps us understand the effectiveness and the topics of content fans find most interesting, and also helps the All England Lawn Tennis Club and Wimbledon content creators proactively, and effectively target their online communications.
So, while Djokovic might be the top seeded men’s player, Serena Williams ruled social influence and authority in the first half of the tournament, abd Roger Federer’s content has the highest predictive impact on fans’ online behavior. Similarly, we looked at who among celebrities frequent Wimbledon, and chat about it online. We found that Niall Horan – a member of the boy band One Direction – leads this year’s social media celebrity sphere by posting just a few messages.
Ohhhhhh @NickKyrgios is at it again! 11 aces in that last set ! Insane young talent
— Niall Horan (@NiallOfficial) July 3, 2015
Brand effectiveness also gave us a bit of a surprise. None of the major athletic brands have appeared in our analysis at the top of the list so far. Instead, this year’s most effective brand in the social space is Evian, provider of the Official Water of The Championships.
We wanted to know why these athletes, celebrities and brands were tops in social media engagement. And what it will mean, in terms of predictions based on what players do online.
In Federer’s case, his influence stems from his timely online fan engagement and also how quickly his fans react to what he posts. Just one tweet from him generate more than 8,000 responses in less than an hour. This voluminous response also makes it possible to make predictions. For example, his post about which court he will play on can cause a reaction that will help us tease out who will be in the audience, such as a celebrity. (Maybe Niall Horan?) It’s also an opportunity for Fed’s brand sponsors to inject themselves into the conversation.
On the other hand, while Serena has a rather large audience (4.75M followers), her authored content is not the major source of the influence. Her impact, instead, stems simply from her vast popularity in being mentioned often, all over the internet. Finally, in comparison to Niall Horan – with his 22.6 millions of devoted followers – he is an example of an influencer with impact achieved through massive amplification. Almost anything he posts post is guaranteed to be amplified from 10 to 100,000 times.
And how about the Evian brand? Evian stands out due to its engaged, although comparatively small community, and due to its active and creative content strategy. From tweeting about celebrity sightings, giveaways, and fun instances of fan interaction at Wimbledon, Evian’s overall audience engagement is more than four times higher than the second-best brand – Stella Artois – and more than all other brands, combined. The bottom line here is that producing creative and innovative content pays off.
The algorithms of influence
Our algorithms, deployed on Softlayer servers, use advanced graph analytics to analyze, quantify, and predict online influence and content effectiveness. We also use Watson Content Analytics to keep spam and noise to a minimum. Paired with our player information, hashtags and other data like location, as well as “influence” algorithms, we can measure fan engagement, and again, predict what might soon trend. If there’s a mention of music or maybe a product connected to an influential player, we can analyze in which geographies, and how quickly the content spreads, and also suggest to the content authors, from marketers to journalists, to perhaps produce more, focused on this emerging trend.
In last year’s Wimbledon a rather interesting set of emerging topics revolved around soccer’s world championship, which occurred at the same time. In some cases soccer-related topics resonated much better that tennis topics – for instance, ironically, Djokovic’s and Federer’s tweets about soccer ranked among the most impactful.
This year, the emerging topics focus on tennis, with the biggest activity spike so far being about the match between Serena Williams and Britain’s hometown star, Heather Watson. Maybe Serena’s “don’t try me” comment had something to do with it.
Some of these insights have been used by media outlets such as Reuter and ESPN. But this use of predictive analysis could transfer across any kind of journalism. Especially now that we’ve woven in geolocation, a reporter could find out what’s resonating across London for story context. We hope to continue testing the technology at future IBM-sponsored live events.