Debates around in-house versus third-party might yet be influenced by the emergence of AI as a significant factor.
Like watching the tide go out and come back in again, there is an ebb and flow to the arguments put forward in the sports-betting arena on which is the best route for success as an operator.
For much of the past five years, a central position has been that building an in-house capability is of the highest priority. Many hundreds of millions of dollars have been spent attempting to prove the on-the-ground reality to back up the theory.
Whether that general trend has been a success will come down to the circumstances of the individual companies involved. But that such moves might not have succeeded in every instance appears to be proved by instances of the countervailing movement towards a greater reliance on third-party systems.
Such would be the conclusion from the news in May that Bally was junking its own, somewhat expensively acquired Bet.works sports-betting platform and switching to the Kambi platform for the U.S. market.
Artificial energy
That deal spoke about the potential for cutting fixed costs, and that obviously remains a driver for any move towards third-party provision. But more than that, the move towards the greater pooling of sports-betting trading and operations under one platform banner also plays into one of the biggest tech trends of the moment.
Paolo Personeni, SVP Managed Betting Services (MBS) at Sportradar, points out that the product journey for his company’s own risk and liability management platform from trading services to a more holistic betting services offering is an instance of how scalability provides benefits to smaller operators that would simply not be available if they were confined to much smaller proprietary systems.
“The evolution of MTS has been driven by our clients,” he says. “It became clear that each bookmaker wanted something different, an offering tailored to their specific needs. We saw an opportunity to develop new products for different market segments, catering for individual bookmakers.”
“Some bookmakers only wanted specific parts of the service, such as risk management on specific leagues and sports. To address this, a selected trading capability was devised, allowing operators to specify the bets they wanted Sportradar to handle.”
As time went on, Personeni says some operators began questioning the need for a full-exposure betting platform and started “considering what core elements could be outsourced”.
The trading platform was one such element that could be entrusted to MTS, allowing operators to interact with the service through it and to focus their attention, energies and resources on other aspects, like player journey, service, marketing, and customer experience.
This shift in approach has enabled smaller sportsbooks to compete not solely based on pricing but by offering superior customer service, a better front-end interface, or more effective marketing strategies.
Simultaneously, the Sportradar betting platform became “increasingly sophisticated”, Personeni says, and this is the point at which artificial intelligence (AI) was introduced. The abundance of data gathered enabled the platform to leverage AI self-learning capabilities and configure itself accordingly.
The platform evolved to be fully powered by AI, leading to enhanced efficiency and effectiveness.
“At the same time, managing everything in-house also became cost-prohibitive for operators,” Personeni says.
“The costs associated with technology, the ongoing need to maintain and improve a trading infrastructure, as well as the need for flexibility and scalability continued to rise,” he adds. “The rapid pace of change, including the challenges of preparing a platform for new jurisdictions, added to the complexity faced by operators.”
The advantages of third-party platforms, Personeni argues, are becoming apparent to ever larger operators and AI is a significant piece of the puzzle.
“There’s great potential to be unlocked with AI,” he says.
In the case of automated player profiling, for example, the technology is being used to analyse and reassess daily a significantly larger volume of player profiles so operators can determine the type of bets they offer to customers based on insights to inform assessments of the risk involved.
“We’re empowering customers to achieve more with fewer resources. We’re providing our partners with tools and services that reduce costs or increase output for customers, while at the same time protecting them from exposure to risk.”
By allowing customers to utilize Sportradar’s services to refine strategies, monitor operations, or collaborate with the trading desk, it leaves the operators able to optimize their trading teams and streamline operations, ultimately increasing the return on their investments.
AI contain multitudes
The MTS product has gone through multiple iterations in the past decade and Personeni says the volume expectations for the product were “consistently exceeded”.
“Looking to the future, the next phase involves the roll out to all sports, as soccer is already in production, of a real-time monitoring dashboard akin to those found in financial markets,” he says. “Within this dashboard there’s a Monte Carlo simulation continually assessing in real time all outstanding and settled bets, providing bookmakers with expected profit distribution and hedging solutions.
It offers a “real-time, AI-assisted financial viewpoint” which Personeni says is an extremely powerful asset for operators looking to make gains in this hyper-competitive marketplace.
To further its ambitions, Sportradar acquired the Optima platform in 2019, recently re-introduced it to the market as ORAKO, as a base to develop and integrate new services. This enabled the company to move towards a modular system, encompassing not only trading services but also other sportsbook functions such as real-time CRM, personalization, and player retention mechanisms.
According to Personeni, the dynamic nature of AI systems allowed for real-time parameter adjustments, potentially running without human intervention, but open to interaction with traders and marketing people who want to finetune and optimize their performances to their own individual goals.
This is assisted by that other significant element of any third-party system, the collection of vast amounts of data. Personeni points out that with Sportradar’s system is transacting approximately 600 million bets monthly, the system’s extensive reach, encompassing various types of bets, multiple currencies, and a global presence, would position Sportradar as “one of the largest and most diversified bookmakers”.
To be clear, size isn’t everything when it comes to sports-betting supply. But in AI, size (as in the sheer amount of data available) does make a difference. Moreover, as third-party sportsbooks generate ever greater data sets, the feedback loop starts to look attractive to other, bigger operators who can also see the evident benefits of scale. This will have an effect on all internal discussions regarding whether to stay in-house or go third-party.
“It feels like AI might be tipping the scales when it comes to questions in this area,” Personeni speculates. That could be highly significant.
