In the spirit of furthering discussion on visual earned media measurement, we are showcasing a Brand Virality Leaderboard ranking the sponsors of the English Premier League. Using a machine learning system from Logotrakr, the leaderboard tracks team sponsor logos in images and video on Twitter and measures the virality of those images against their peers. Think of it as a Nielsen Rating for BPL sponsors.
You’ll see some expected names at the top of the table, but also some very surprising up and comers and a lot of back and forth each day.
Check out the leaderboard here.
Sponsors are ranked by virality for the past seven days. Scroll down the list and mouse over the position arrows to see the rank change each day. The data presents a rolling weighted score of tweets, retweets, favorites and audience for the previous seven days. Each brand is given a score that reflects the virality of social activity around media that contained their logo found on Twitter. The board updates at midnight GMT each day.
Technology is changing football! The SportTechie newsletter keeps you in the know.
To normalize the data, images and video are collected from hashtags only, without regard for @users or @mentions, with a bias towards persistent hashtags such as #BPL rather than game or player hashtags that can introduce extreme volatility in the numbers. The goal is to give an objective look at the visual reach of each sponsor on Twitter and spark more debate on how best to use this technology in the future.
The benefits of visual tracking of sponsor logos are several. Marketing teams at clubs can compare themselves directly to their peers and develop best practices to improve their reach and engagement scores. Brands can independently verify the efficacy of their sponsorships and directly engage in conversations with the audience. Agencies can tangibly measure social media ROI for their clients and make strategic recommendations on sponsorship based on hard data.
Images and videos posted to social media that include a sponsor logo in some way, such as on a jersey or signage, are creating value but no one agrees on how quantify it. Some of these images are shared thousands of times and have millions of potential viewers without an agreed upon standard for measuring the impact and currency value of these engagements.
It is well established that visual content is far more engaging for users than text only content. Clearly any future measure of earned media on social networks must include a visual component to be meaningful.
The idea that brands get “a huge amount” of visual exposure from their sponsorships via social networks is generally accepted. The rationale for using machine learning to detect the visual impact of sports sponsorship on social media is to apply the same sort of specificity to visual measurement as there is in text on the internet.
This is particularly true for the elusive engagement piece of the puzzle, where the brand is literally underwriting the social conversation through their sponsorship of the club jersey and yet has no idea that the conversation even happened. Here is a classic example of a tweet with a huge potential audience that the sponsor is likely unaware of:
Schweinsteiger: “I face a new challenge every day at @ManUtd. That helps me develop further.” #UCL #MUFC pic.twitter.com/Kja4nPFxM5
— Champions League (@ChampionsLeague) September 1, 2015
So far the conversation around visual earned media measurement has focused on case studies or specific events with a small number of brands. Given the interest in this topic and the large audiences at stake, a real time demonstration of what the machine learning can offer over a large number of brands across an entire league seems particularly relevant.
A logical starting place for a practical application of this technology is the Barclays premier league. The BPL has a global and growing audience, a valuable sponsorship market, and active social media staff teams. This combination of audience, activity and intent is the perfect proving ground for comparing brand sponsorship effectiveness.
In the days ahead, we will share examples of the stories behind the leaderboard, identify trends and best practices for using this technology to measure sponsorship ROI.