Adam, the traditional role of the platform has been to provide the infrastructure connecting an operator’s products and systems. How significantly is that role changing as operators expect platforms to make decisions as well as move data?
The shift has been significant. Historically, platforms operated as passive link technology, moving data from one point to another, storing player records, and operating various other tasks.
Operators today operate in an environment of sheer, real-time volume of player events, with thousands collected every second. They can no longer afford the latency of extracting data, analysing it here, and then pushing a decision there. The platform’s role is now to evolve from the passive actioner to the live decision-maker. It has to transport the data while interpreting signals as they fire.
From Xtremepush’s perspective, where should the intelligence behind the customer journey actually reside – within the core platform, within specialist technologies such as CRM and engagement systems, or across both?
Intelligence should reside where the execution happens to avoid any latency, which points directly to specialist engagement systems like CRM and loyalty tools in our sector. However, for this to work effectively, the core platform must provide a robust, unified data layer.
The more silos and link points that go untreated with what we believe to be best practice with this aligned, always-on data architecture, the more inefficiencies arise, the more money gets left on the table, and the weaker the relationship comes between the operator and its end user.
Operators now have access to huge volumes of real-time player data. What separates simply collecting and centralising that information from turning it into useful decisions and actions?
Many operators have successfully centralised their data, but they still process it in batches. The separator is that real-time nature that allows player activity to be processed and acted upon in real-time, with automated frameworks and seamless data flows.
Players today are used to live interaction with entertainment platforms, whether it’s a recommendation on Amazon or a next watch on Netflix, and there’s no reason that casino players or sportsbook bettors should receive responses to their own behaviours 12 hours later.
We have to treat every bet, click, and deposit as a live trigger, and when paired with engagement engines that plug into quality platforms, only can we then deploy relevant content based on those micro-moments.
How is AI changing the way operators determine what message, promotion, game or experience should be presented to an individual player – and, importantly, when nothing should be presented at all?
AI is moving us away from rigid segmentation and calendar-based campaigns, and towards agentic, hyper-personalised engagement. With tools like Xtremepush’s XpertOS, we can deploy autonomous agents that analyse millions of behavioural data points to calculate the commercial value of engaging a specific player at a specific time. This goes far beyond what any CRM team could achieve, whatever the size. The key benefit is that underserved players and commercial opportunities are realised faster.
Real-time decision-making depends on information moving rapidly between multiple systems. What does the underlying platform architecture need to look like for that intelligence to work effectively?
It requires a shift away from fragmented, API-heavy “bolt-on” architecture towards native data layers. Every time data hops between a disconnected wallet, a standalone gamification tool, and a legacy CRM, you introduce latency and the risk of incomplete information.
The underlying platform must be built on event-driven streaming, where state changes are broadcast by all connected technologies. The CRM and loyalty engines must sit directly on top of this real-time stream, executing logic or planned actions in milliseconds rather than casing a database every few minutes.
Personalisation is often discussed in terms of better targeting, but can an intelligent platform also identify when an operator should intervene differently – for example because a player’s behaviour, value or engagement pattern has changed?
This should be a prerequisite. Whatever the amount of data at hand, a marketer can only make so many decisions in a day, and much of that would be educated guesswork too. The right platforms feature richer engagement engines that allow for personalisation on an almost one-to-one level.
When we can interpret those signals faster and with more accuracy, we can then serve players more effectively, therefore improving their user experience and realising greater lifetime value to operator customers. This process may be a simple flag of a certain type of golf bettor, for example, that has been underserved despite activity picking up in an appropriate manner over time, or it could be toning down messaging to a casino cohort that has shown potentially risky behaviours on a certain type of game format they’ve just come across.
If several connected technologies are capable of making decisions about the same customer, who should ultimately determine the next action? Does the industry need a clearer hierarchy or orchestration layer for real-time decision-making?
Operators need a centralised orchestration layer, and logically, the CRM should serve this function. The CRM is the only system designed to view the individual player holistically across all lifecycle stages. If your risk engine flags a behavioural shift, your loyalty system triggers a level upgrade, and your sports platform wants to offer a bet boost, all of these signals should funnel into the CRM.
It’s best suited to weigh those priorities, apply any compliance guardrails, and determine the optimal action to take.
Xtremepush works with operators and platform providers through integrations and partnerships. What are operators now asking for that they were not asking for two or three years ago, particularly around real-time data, automation and AI?
A couple of years ago, conversations were dominated by channel expansion – adding WhatsApp, for example. Today, operators are asking for autonomous execution. They realise their human teams cannot manually build campaigns for the hundreds of micro-segments still to be discovered in their data, and so they are asking for solutions that realise these commercial goldmines, that also face-off increasing margin pressures and regulatory squeezes.
The development of XpertOS begun two years ago ahead of this blanket requirement, and it’s now live, ready to serve those who are looking to incorporate the future of CRM technology today.
Looking at the direction of travel, does the concept of the “platform” itself need redefining? Are we moving towards an architecture where its primary value is no longer simply providing infrastructure, but coordinating the intelligence running across the entire operation?
In some aspects, absolutely. The “platform” of the future is not passive. It is an active operating system for customer engagement in our case. As AI scales, the platform’s primary value will be acting as the governed environment where real-time intelligence operates safely. It can provide the guardrails, such as compliance checks and suppression lists, ensuring that autonomous agents can operate effectively with no associated risk from delegation. This environment is going to shift rapidly in six months, two years, and beyond, but what will remain critical to a successful CRM platform is real-time data, agentic insight and execution, and governance as a priority throughout.
Adam, the traditional role of the platform has been to provide the infrastructure connecting an operator’s products and systems. How significantly is that role changing as operators expect platforms to make decisions as well as move data? The shift has been significant. Historically, platforms operated as passive link technology, moving data from one point to