Why iGaming operators are still getting real-time data wrong
As iGaming operators continue investing in artificial intelligence (AI), customer relationship management (CRM) automation, and player engagement technologies, the industry is placing increasing focus on how data can be used to improve acquisition, retention, and long-term player value. Across the sector, operators are moving towards more personalised and predictive engagement strategies, while also facing challenges linked to fragmented systems, data latency, rising infrastructure costs, and growing regulatory scrutiny around player protection and responsible gaming.
In an exclusive conversation with SiGMA News, Christoffer Feldt-Sørensen, Chief Sales Officer (CSO) at Symplify, a Sweden-based customer engagement and marketing automation platform, explains why many iGaming operators still misunderstand “real-time” data and why continuous data streaming is becoming more important for player engagement and retention. The interview also looks at artificial intelligence-driven retention tools, predictive churn prevention, behavioural data, bonus-led acquisition strategies, and the role of responsible gaming in data-driven operations.
SiGMA News: Everyone in iGaming claims to use real-time data. What are operators actually getting wrong when they make that claim in practice?
Christoffer Feldt-Sørensen, CSO at Symplify: “Real-time data” has been misused for so long that operators have lost trust in the term entirely, and rightly so. What most vendors call real-time is actually near real-time, with latency built in. The focus is now shifting towards data streaming, where transactional data moves instantly rather than in delayed batches. Think of it like Netflix. If your connection is working, the movie just plays. No buffering. That means player interactions can be triggered the moment something happens, not seconds or minutes later. The distinction sounds small, but the commercial impact is significant.
SiGMA News: As operators invest more in data infrastructure, expectations around real-time capabilities are rising. What should they prioritise next to make these systems actually deliver value?
Feldt-Sørensen: The infrastructure investment means nothing if the CRM team still has to raise a ticket and wait a week every time they want to build a new segment. That dependency on tech is where value is lost. Operators need to put more control directly in the hands of the people closest to the players. A CRM team should be able to define a segment in a morning meeting and have it live the same day without technical involvement.
Once that is in place, operators can move from broad player journeys to highly specific ones, sometimes involving just five to seven players. That might sound counterintuitive until you realise those are often VIPs and high-value players, where retention and engagement can have the greatest commercial impact.
SiGMA News: AI and real-time data are often positioned as industry solutions. Where is the biggest gap between what AI promises and what it actually delivers?
Feldt-Sørensen: The gap is not in the technology. It is almost always in the foundation underneath it. Tech teams come in confident, saying they have everything covered, and then run into problems when reality hits. Churn prevention tools are a good example. Instead of waiting for a 30-day retrospective, they work day by day, identifying players likely to churn far earlier than manual analysis would.
However, in the early stages, the process is slower because the system is still learning. Results improve over time rather than overnight. Operators who become disappointed with AI are often those who skipped the groundwork, such as dealing with messy data, disconnected systems, and unclear strategies, while expecting the technology to fix everything on its own. AI amplifies what is already there. If the foundation is weak, AI simply accelerates the problem.
SiGMA News: Most acquisition strategies still depend heavily on bonuses and promotions. What would a genuinely data-led acquisition model look like without relying on bonuses?
Feldt-Sørensen: The bonus-led acquisition model is a bit of a trap. It is easy, it converts, and the numbers look good on day one, but operators are largely acquiring players who are there for the bonus rather than the brand. When the bonus runs out, so do they, and operators are back to square one with the acquisition budget.
A genuinely data-led approach starts by asking a different question: not “How do we get players to sign up?” but “What kind of players do we actually want, and where do we find them?” That means using behavioural and demographic data to build a profile of valuable long-term players and then shaping the acquisition strategy around attracting more of them through relevant content, the right channels, and messaging that reflects what they actually value.
The CRM team and marketing team need to work from the same data and the same goals from day one, rather than handing off a converted sign-up and hoping for the best. Acquisition is only as strong as what happens next.
SiGMA News: Can you share one situation where real-time or data-driven engagement clearly improves retention, and one where it leads to a poorer player experience?
Feldt-Sørensen: A positive example would be a player hitting 18 consecutive losses on a slot. Using data streaming, a personalised pop-up can trigger immediately, offering five free spins. The player continues, the session extends, gross gaming revenue (GGR) increases, and the player feels the brand understands them. That is an example of real-time engagement working effectively.
A negative example is over-automation across channels. Many players receive the same message by email and then again through a push notification moments later. When automation is not governed by clear cadence rules and channel preference logic, it becomes noise rather than engagement. The player feels harassed rather than valued. The responsibility should not fall on the player to correct that experience themselves.
SiGMA News: Operators collect large volumes of player data across multiple touchpoints. What is the biggest gap between the data they collect and the data they actually use effectively?
Feldt-Sørensen: Operators are generally strong when it comes to transactional data such as bets, wins, and losses because the connection to revenue is obvious. However, the behavioural layer sitting on top of that remains largely untapped. What is the player looking at before making a deposit? Where are they dropping off? What do their browsing patterns reveal about what they actually want to play? That data already exists, but much of it remains unused.
AI-driven game recommendation tools are increasingly being used to present players with games tailored to their behaviour, and the impact on GGR can be significant. Operators that are not using this type of data are not struggling because their transactional data is inaccurate. They are struggling because they continue to treat every player the same once they are on the site.
SiGMA News: Many platforms position themselves as predictive and AI-driven. What capability must operators have before predictive and AI-driven systems can work?
Feldt-Sørensen: Streaming data. Not near real-time or batch processed data, but actual streaming. Everything else is built on top of that foundation. If data arrives with latency, AI systems are making predictions using outdated information, and triggers are firing too late to have an impact.
Another important factor, which is often overlooked, is that the CRM team needs the autonomy to act on what the AI identifies without going through a technical bottleneck. Bonus abuse detection models, for example, can flag a player within the first five minutes of registration and automatically remove bonus communication from that player’s journey. That only works if the system, the data, and the team are aligned and operating at the same speed.
SiGMA News: At what point does real-time or data-driven execution start to break down for operators, whether due to cost, complexity, or infrastructure limitations?
Feldt-Sørensen: There are two main points where it begins to break down. The first is cost. Streaming transactional data at scale comes with a significant price tag, and operators that have not yet seen the return are often reluctant to make the investment.
The second is internal knowledge gaps. Tech teams often overestimate their readiness, run into difficulties during implementation, and stall the process. A phased approach is generally more effective, starting with basic flows, building trust in the technology, and then gradually adding complexity. Operators that attempt to scale everything on day one often struggle.
SiGMA News: Data-driven strategies are often linked to revenue metrics such as GGR and net gaming revenue (NGR). Where are operators overestimating the commercial impact of these tools?
Feldt-Sørensen: Operators often overestimate the impact when they buy a platform and assume the work is done. The operators genuinely improving their GGR and NGR figures are the ones investing time in understanding their players, not just their data. There is a difference. Data tells you what happened. Understanding tells you why it happened and what to do next.
Operators that struggle are typically chasing quick results, running a small number of broad player journeys, and wondering why performance is not improving. In contrast, operators that spend time refining segments, becoming more granular, and allowing smaller journeys to run over longer periods are often the ones seeing stronger retention and more meaningful player engagement.
SiGMA News: There is growing pressure around responsible gaming and data privacy. Can data-driven personalisation genuinely support player protection, or is it still primarily used to drive revenue?
Feldt-Sørensen: Data-driven personalisation can support both player protection and commercial performance. Operators that treat those goals as opposites are missing the point. The same data streaming used to trigger a free spin offer at the right moment can also identify signs of risk, such as a sudden spike in deposits, sessions running far beyond a player’s normal pattern, or behaviour that significantly differs from previous activity.
When those indicators appear, promotional communication can be paused automatically and the player can instead be directed into a responsible gaming journey. Compliance is becoming increasingly important across regulated markets, and operators that rely on personalisation purely for short-term revenue growth are likely to face greater regulatory scrutiny. A more sustainable approach is building systems where player protection and commercial performance operate together.
SiGMA News: From your perspective at Symplify, with visibility across multiple operators, what is one industry habit around data or CRM that needs to be unlearned or fundamentally changed?
Feldt-Sørensen: Retrospective thinking. The vast majority of operators still carry out churn analysis by looking back 30 days and reacting to players who have already left. That is not retention. It is rescue.
The habit that needs to change is waiting for signals that are already obvious. Operators leading the industry are increasingly using AI to identify churn risks days or weeks before they happen and intervening while there is still a relationship to preserve. Moving from reactive to predictive systems is not only a technology upgrade. It is also a broader shift in how operators think about player value.
Subscribe HERE to SiGMA’s Top 10 News countdown and SiGMA’s weekly newsletter to stay up to date with all the latest iGaming News from the biggest iGaming community in the world and benefit from subscriber-only offers.
