It’s Wednesday morning. Work sucks. You decide that what you really need to get you through your midday meeting is some magic from your favorite footballer. So you log onto Liverpool FC’s TV GO platform, search for ‘Mohamed Salah crushing Man U’ and, instantly, there it is: a 30-second compilation of the Egyptian King’s goals against the Red Devils. A perfect shot of digital caffeine, created just for you.
This type of use case is exactly what the top team at Liverpool FC are focusing on as they “think really carefully about where, how, when we’re applying AI”, according to the club’s senior vice president of digital, Drew Crisp. The instantaneous stitching-together of personalized content on demand may be less headline-grabbing than reports of AI being used to perfect player performance, such as in LFC’s own TacticsAI project with Google DeepMind, but “the ripe hunting ground for AI for us is in the media space,” Crisp insists. “It’s such a massive opportunity not just from an efficiency angle but in terms of speed, cost reduction, capacity. It enables us to reach fans in a different way.”
Putting fans first
Other industry leaders agree that, despite all the fuss about AI being used in coaching, transfers and recruitment, fan engagement’s where it’s really at when it comes to the technology’s potential to drive revenue for clubs. Alexandra Willis, former communications and marketing director at Wimbledon, and now director of digital media and audience development at the Premier League, has focused on tech innovations to enhance fan engagement throughout her career, including the award-winning deployment of first-of-a-kind AI highlights in partnership with IBM at Wimbledon.
“There are many established use cases for AI within the sphere of sports marketing and content creation, which are only going to accelerate as large language models become more sophisticated,” she says. “Many of these use cases – deployed by the likes of the NBA, Wimbledon, Augusta and across football, have focused on delivering increases in scale and efficiency of production through multi-variant AI-generated highlights, summarizations, and commentaries. These are the foundations for hyper-personalization and hyper-localization, which are important fan engagement trends.”
LFC is working with Boston-based Wasabi Technologies as it attempts to meet these trends. Launched two weeks ago at the National Association of Broadcasters trade show in Las Vegas, Wasabi’s AiR service uses AI to scan, tag, translate, and compile content from clients’ cloud archives in a move that Wasabi’s CMO, Michael Welts, calls “the most significant advance in the storage industry since the invention of object storage.”
“Wasabi AiR ingests video archives directly to Wasabi hot cloud storage,” he explains. “There, Wasabi AiR analyzes the data and creates second-by-second metadata that allows users to quickly identify and surface the video content they need. This allows sports, media and entertainment organizations like Liverpool FC to upload nearly a century’s worth of video content for analysis, and make formerly ‘dead’ content a new source of revenue and user engagement.”
Content 800 million ways
The ability to automatically tag and tailor content in myriad ways is especially relevant for football, which crosses borders like no other sport. As the most-watched team in the world with over 800 million viewers globally, LFC could benefit hugely from adapting match footage to different markets; think foregrounding players popular in the States, or picking out key moments that data predicts will appeal to viewers in Asia, location of the club’s largest global fanbase. But one experiment Crisp’s team definitely won’t be taking public – yet – is the digital manipulation of its stars into multi-linguists, however big the potential rewards.
“We’ve trialed some of the generative AI capabilities to take content from players and managers and put it into a local language; not just dubbing, changing facial expressions and movement too. [But] we can’t just post that without being impeccably clean, impeccably honest and very respectful of people’s personal image and rights. We always want to do things the right way, so we are treading carefully.”
Willis agrees. “We have to give as much focus on the trust, credibility and data protection aspects of AI as we do to its generative possibilities,” she says. And one knock-on advantage to Crisp’s caution is a high level of buy-in from his team. “I think when producers and editors can see how it gives them time back to do the fun stuff as opposed to ‘I’ve got to find 73 bits of content to make this work’, actually they then really embrace it, because they can see how it helps them, rather than looking like they’re going to be replaced.”
An open field
The need for impeccable cleanliness may be one reason that AI adoption in other Premier League clubs is patchy at best; at least, as far as they’re willing to publicly let on. Chelsea is an outlier with its Match View FX feature, a platform which sits inside the Chelsea app providing highlights, breakdowns and gameplay data on top of each broadcast. In contrast, Arsenal’s most recently publicized project dates from 2018, in the form of the Robot Pires chatbot, created with the AI chat app GameOn. And although Liverpool’s historic rivals Manchester United announced a new AI partnership this January with Manchester Metropolitan University’s Institute of Sport, the focus is once again on performance analysis rather than fan experience.
Crisp is positive, however, that this is the start of a powerful cross-industry transformation, for football execs and fans alike. “We’ve gone beyond discussion and we’ve gone beyond PowerPoint slides and people telling us how cool AI is. We are finding a few areas where we can actually do something meaningful, and in the future we plan to explore many different applications of AI across a multitude of categories with the right partners. So what we’re doing right now is early stages, but nevertheless it’s real, and proving results.”
