From niche to new vertical: The evolution of prediction markets
Prediction markets have exploded in popularity in a short space of time, transforming rapidly from a niche concept into one of the most talked-about developments in betting.
A major catalyst was the US election of late 2024, which brought prediction markets into the mainstream. Since then, growth has accelerated rapidly.
According to H2 Gambling Capital, prediction markets handle for March Madness reached $16m in 2025. A year later, the figure for the college basketball tournament was projected to rocket to an estimated $530m.
Rapid rise
Yurii Berest, CEO of sportsbook solution supplier DATA.BET, believes there are several factors contributing to the rapid rise of prediction markets – including a broader range of events people can engage with, and how those markets are being presented.
“Historically, betting has been concentrated around sports,” Berest says. “Prediction markets extend that behaviour into areas people already follow closely in everyday life, including politics, technology, economics, entertainment and other major real-world events.
“For many users, those subjects are at least as familiar as traditional sports markets, which significantly broadens the potential audience.
“At the same time, prediction markets have become easier to understand and more engaging as products. Users are not only betting on an outcome; they can also follow how the market’s implied probability changes as new information emerges. That makes the market itself a form of real-time content.
“I think the combination of a much broader range of relevant events and a more accessible way of interacting with them is what has pushed prediction markets from a relatively niche format into the mainstream conversation.”
Presentation and accessibility
The importance of presentation and accessibility was central to DATA.BET’s launch of its own prediction markets offering in April.
A “sportsbook logic” was harnessed, with the company aiming to reduce the structural complexity of exchanges by introducing clear pricing, fiat transactions and an intuitive user experience.
Operationally, it’s a new vertical; experientially, it’s designed to feel like a natural extension
However, prediction markets are firmly a standalone vertical within the DATA.BET solution, with players able to bet on events covering everything from politics and finance to culture and the weather without navigating complex mechanics or using cryptocurrency.
“We see it as a separate category, even though it sits within the sportsbook experience,” Berest says. “On the surface, users interact with it through the same UX they already know. From the operator’s perspective, integration is possible via the API, so no new additional integrations are required.
“Underneath, it required building an entirely new subsystem — different pricing logic, different risk management, odds calculated dynamically based on incoming stakes and fast-changing data. So operationally, it’s a new vertical; experientially, it’s designed to feel like a natural extension of the familiar sportsbook that DATA.BET provides to its clients.”
A familiar betting interface
DATA.BET’s offering, which is available via iFrame and API integration, has been built specifically for sportsbook users rather than traders. This means players have a familiar betting interface and can use standard decimal odds, their existing account and the same balance they already use with the operator, whether in fiat or crypto.
Since the initial vertical launch, DATA.BET has introduced additional market types and richer live and historical market data. This chimes with a broader trend at the heart of the growth of prediction markets, as reflected by the rise of platforms such as Polymarket.
Prediction markets are a good example of how a mature industry can still create new engagement
“I think the biggest lesson is that the betting product itself still has significant room to evolve,” Berest explains. “The growth of platforms like Polymarket shows that users are interested not only in traditional sports betting, but also in new types of content and new ways of engaging with events they already follow closely.
“It also shows that the market itself can become part of the experience. Users can follow changes in probability, volume and market sentiment even before they decide to place a bet.”
New formats, mechanics and categories
For the iGaming industry, though, the lesson is not necessarily to copy the exchange model, Berest adds. Instead, the focus should be on looking for new formats, mechanics and categories that are relevant to users, while adapting them to an experience that is simple and familiar.
“Prediction markets are a good example of how a mature industry can still create new engagement by rethinking what users can bet on and how they interact with that content,” he says.
One of the most significant differences between prediction market products lies in the design of their underlying models, which range from peer-to-peer exchanges and liquidity-driven pricing to fixed-odds approaches.
“Each model has its strengths,” Berest explains. “Exchange-based markets are strong at price discovery because prices reflect market activity, available liquidity and the behaviour of participants. They can also make users feel that they are participating in a market rather than simply betting against a bookmaker.
“The trade-off is complexity. Order books, liquidity depth, partial fills and trading-style execution can be intuitive for experienced traders, but they create additional friction for a typical sportsbook user.”
DATA.BET combines elements of both models. While the user experience is deliberately familiar, the pricing retains market-driven mechanics, based on information from exchanges, available liquidity and calculations based on volume-weighted average price (VWAP).
“This means users do not need to interact with an order book or learn a trading interface, while the pricing still responds to liquidity and market activity,” Berest says. “For operators, it provides a way to introduce prediction markets without forcing their existing sportsbook audience into a completely different product experience.”
Data, AI and automation
Real-time data and automation play a fundamental role in making prediction markets work at scale – and not just in relation to market-driven pricing.
“Markets can move quickly, so pricing, exposure management, market suspension and settlement workflows need to operate with minimal manual intervention,” Berest adds.
Users need to understand how a market will be resolved and have confidence that the pricing process is consistent
“Event monitoring is also becoming increasingly important, particularly for markets tied to fast-moving real-world events. The system needs to identify when conditions have materially changed, when a market should be suspended and when the predefined resolution criteria have been met.
“AI can support some of these processes, especially around monitoring, classification, anomaly detection and operational automation, but the core pricing and settlement logic still needs to be transparent, deterministic and based on clearly defined market data and resolution rules.”
Instilling long-term trust
Clear rules and consistent settlement will be crucial to building long-term trust in prediction markets.
This means settlement rules must be defined before the market opens, removing any room for interpretation. Settlement should then follow those predefined rules through an automated and auditable process, rather than discretionary decisions after the event, Berest says.
“Users need to understand how a market will be resolved and have confidence that the pricing process is consistent,” he adds.
We expect to see more short-duration and frequently resolving markets, creating a much more repeatable pattern of engagement
“Pricing transparency is also important. Users do not necessarily need to see the full internal pricing model, but they should be able to understand that prices reflect real market activity and liquidity. Providing volume data and probability history helps users see how the market has evolved over time.
“Ultimately, consistency is what builds trust. The same rules need to be applied in the same way across markets, regardless of the outcome or the operator’s exposure.”
Repeatable engagement patterns
Looking ahead, Berest believes prediction markets will gradually become less defined by the label itself. Instead, the more important shift is that real-world events are becoming another category of “bettable” content alongside sports, esports and other established verticals.
“As the category matures, we expect to see more short-duration and frequently resolving markets, creating a much more repeatable pattern of engagement,” he says, before adding that the user experience will also continue to become simpler.
“Most users will not care whether the underlying pricing comes from an exchange, a liquidity model or a sportsbook trading system,” he says. “They will care about whether the market is easy to understand, whether the price is competitive and whether they can place a bet immediately using the account and balance they already have.
“That is where I see the category heading: broader real-world content, faster market cycles and increasingly sophisticated market mechanics operating behind a much simpler user experience.”