Tech focus: AI and data analytics deliver deeper insight into sport

AI is expected to drive continued spend on sports media technology in two ways, according to Tom Morrod, research director and co-founder of Caretta Research. First, by making tools more accessible to a wider range of buyers, particularly sports clubs, leagues and federations, and secondly by creating more data from sports, games and matches by using AI rather than manual logging.

Morrod says the GenAI hype has finally peaked this year, with generative types of AI much less widely used than machine learning algorithms in this field. He highlights that while AI might be limited in the value-creation tasks it can achieve, there are lots of cost reduction tasks it can perform to free up people to discover new value creation elsewhere.

“Data analytics used to be a complex and specialist task but now we are seeing tools that get better at suggesting how to use data, spotting interesting trends in the data, creating views and dashboards, helping things run better in a variety of manners, from content-aware encoding to chatbots. Ultimately the clever stuff comes from what the people do with these capabilities and the free time they save.”

Caretta Research data on sports analytics shows growth from about $400 million in 2018 to almost $900 million at the end of 2024, much of which is driven by sports buyers directly, rather than the traditional buyers in broadcast and pay TV.

Indeed, AI and data analytics are enabling the sports industry to unlock unprecedented storytelling potential, according to Samir Gole, VP of product at sports AI company Stats Perform. The company has combined the latest AI capabilities with its 6.5 Petabytes of proprietary Opta sports data to help clients and partners like NBC, BBC, Sky Sports, Paramount and Vodafone to perform deeper analysis, find stories faster, deliver real-time data-led storytelling and create new commercial opportunities.

“New AI-enriched data sets make it possible to extract never-seen-before insights about player performance and team strategies and quickly surface attention-grabbing details about a game so that sports commentators, analysts, researchers, and journalists can tell more engaging stories,” explains Gole. “The ability to explain the stories behind every moment of a game is now a non-negotiable requirement for all our customers. Tools like Opta Stream, Opta Stories and the recently launched OptaAI Studio also increase sponsorable inventory, connect to ad networks, and provide rich digital and broadcast sports experiences, before during and after games.”

Next-level fandom

The rapid proliferation of AI in sports has given unprecedented access to data for coaches, scouts, analysts, and fans. Dave Greene, sales director at Studio Automated, believes this has created a new type of fan – the ‘prosumer’ fan – that engages sports not just by watching the event but also by understanding the ebbs and flows of the match via data and statistics made available, sometimes in combination with ancillary experiences via gamification, betting, and fantasy sports.

“GenAI is at the forefront of personalisation. Broadcasters, OTTs, and rights holders are only just coming to terms with the capabilities afforded to them, in my opinion,” says Greene. “I believe that in the future you will see more engaging content driven by brands and sponsors via Generative AI. Content can be tailored to audience segments in a highly targeted way to create impactful and unique campaigns that further enhance the connections between a fan, a player or team, and the brands that invest in the sport.”

“The ability to explain the stories behind every moment of a game is now a non-negotiable requirement for all our customers”

Studio Automated has developed an AI-automated controller for PTZ cameras to help deliver high-quality video resolution, and AI-automated panoramic camera solutions to provide real-time footage to coaches and analysts on the touchline to quickly review the action from an alternative angle so they may make informed strategic decisions. Additionally, the company offers event detection, for coaches and for OTT and rights holder partners.

Beyond data: Mining audio and video

An AI company since 2014, Veritone specialises in enterprise AI software, applications, and services around audio and video. The company helps customers in sports and media entertainment make content easier to search and discover. This makes it easier to perform sponsorship verifications and to monetise that content for secondary and tertiary use cases.

Veritone is seeing increasing numbers of organisations treat their video and audio assets like they would traditional structured data, which allows them to mine the full range of their intellectual property and the assets they create.

“We ingest hundreds of thousands of hours of audio a day, append and annotate the data analytics around it in near real-time: who’s on the screen, what are they doing, what sport, what are the numbers on the jersey, what logos are in the background and so on,” explains Sean King, GM of Veritone Media and Entertainment. “AI is making it smoother and easier to manage these volumes and allows us not only to pull data around an event, but also the exact moment in the video and game it occurs and being able to service that.

“We’ve done a lot of work with customers like the San Francisco Giants that started to dive into their entire content archive to improve the fan experience. To engage with fans, the marketing team needs to find those moments – many of which were trapped within an analogue archive of different types of content – and build that nostalgia. It’s about being better able to utilise what you have, tell stories, bring memories back and keep fans engaged in the stadium and in the off season. These are volumes that cannot be handled by a human.”

AI Media, specialising in AI-generated real-time speech to text, producing subtitles both in the same language and translation, also highlights the benefits of being able to work with AI beyond core speech to text recognition, by sitting within the customer workflow to leverage visual data and context.

“Using AI, we filter out noise to produce subtitle text with sub-one second latency and translate the commentary,” explains John Peck, VP international sales at AI Media. “For sports, beyond reaching viewers who are deaf or hard of hearing it’s about eyeballs and monetising content. AI translation tools with subtitles allow stakeholders to reach audiences that they may not have reached with that content before.”

AI Media is launching an AI voice product called LEXI voice in April 2025, which builds on this. “Rather than the end-result being translated text subtitles, it will be a translated audio track. This will allow the distribution of content to a completely different region bearing not only local subtitling, but also an audio track in the local language using a synthetic voice that carries emotion as well. You now have a very inexpensive way to reach and grow a new audience without having to find local resources. It’s a low price, low risk investment for potentially big returns.”

Peck also sees some interesting potential in feeding large language models (LLMs) their favourite diet: vast quantities of text. “We have access to all of this text produced from audio for the subtitling and translation services in a sub one second timeline and we think we can leverage that to provide insights to people in the sports broadcast chain.”

Subscribe and Get SVG Europe Newsletters