Matthew Starker of Endeavor Streaming provides practical examples of how streaming services can utilise data to maintain sports fans’ interest
OTT streaming is a key part of the future of global sports delivery. We’ve seen Prime Video claim the rights to Thursday night NFL games, Apple TV+ score deals with the MLB and MLS, Max launch a live sports tier—the list goes on.
Many sports federations and teams have also launched owned-and-operated OTT destinations, such as Real Madrid’s RM Play or NBA League Pass, where they can create true superfan paradises with libraries of content outside of game day.
While these platforms have no problem finding an audience and capturing fans’ attention during live events, the challenge lies in maintaining engagement and preventing churn year-round. Enter machine learning (ML) predictive analytics.
Predictive ML models output scores that quantify a user’s propensity to commit a certain action; for example, their likelihood to churn, stream content, or buy pay-per-view. They provide an avenue for platforms to proactively action data to mitigate subscriber loss. Two of the strongest use cases lie in targeted content recommendations and marketing communications.
Targeted Content Recommendations
Due to the sheer amount of content available for fans’ consumption, discoverability via personalised content recommendations is central to the success of every streaming platform. Sports organisations need to keep fans hungry for content beyond game day. By surfacing the right content at the right time, they can create a tailored superfan experience, foster habitual usage, and reduce churn.
Recommendation engines built on predictive models determine a fan’s likelihood to stream a certain piece of content in the future. For example, if a fan tunes into Prime Video each Thursday to watch the NFL, there’s a high likelihood they’ll be interested in content like the documentary Kelce.
If a fan has watched every Conor McGregor fight on UFC Fight Pass, there’s a good chance they’ll pay to watch the next one. Armed with this information, predictive models can proactively serve up new or undiscovered content—including interviews, locker room footage, or docuseries—that’s hyper-personalised to that fan’s preferences, therefore boosting engagement.
The seasonal nature of many sports lends itself to big churn moments when seasons end or teams are eliminated. Therefore, it’s critical for platforms to know when a fan is likely to churn so they can reach them before it’s too late. Predictive ML models take marketing a step further by helping action usage data so sports entities can be more strategic in their messaging to specific audience segments.
For example, if an ML model can predict a cohort of fans is likely to churn once their alma mater is eliminated from March Madness, the platform can communicate to those fans differently than it would to fans who are unlikely to churn: it could offer a complimentary month of service, a discount on an annual subscription, or perks for live games and merchandise. This provides a chance to re-engage users and re-establish the platform’s value. Plus, by only communicating these offers to those most likely to churn, the organisation saves revenue it may have lost by offering that financial incentive to users who weren’t at risk.
A Data-Driven Future
In the dynamic landscape of sports streaming, ML predictive analytics are a game-changer today, but will soon be table stakes. By leveraging these cutting-edge models to tailor recommendation algorithms and strategic marketing, sports organisations deliver a personalised, best-in-class experience that fosters fan loyalty and combats churn. The future of sports streaming is data-driven, and these models will continue to increase the value of that data and help sports entities stay better connected with their consumers.
Matthew Starker is chief business officer of Endeavor Streaming