How should iGaming operators build a CRM strategy?: Retention, AI and VIP management

Is your CRM driving revenue, or just generating communications? Are you treating deposit as the finish line, when the real engagement opportunity starts after the money is in?

In episode 18 of Connected with Pragmatic Solutions, CEO Ashley Lang speaks with Brigitta Szenczy, Co-Founder of Kyborg, a specialist CRM and player lifecycle agency working with iGaming operators across multiple markets. The conversation covers the structural mistakes operators make in CRM strategy, how to set goals correctly, when to personalise versus when to keep it broad, and how to manage VIP relationships before churn becomes visible.

Deposit is not the finish line

One of the clearest insights in this conversation is also one of the most common blind spots in iGaming CRM: operators build campaigns to drive the deposit, measure success on that deposit, and then effectively release the player. The real engagement decision happens after the deposit, when the player's experience determines whether they return.

This means that CRM strategies anchored solely to conversion and deposit metrics are optimising for an intermediate step, not for long-term player value. Operators who treat the deposit as the conclusion of a campaign are missing the part of the journey where loyalty is actually built or lost.

Strategy before tooling, and before KPIs

A recurring theme from Brigitta Szenczy's experience is that operators too often define CRM goals after the system is already live, or worse, let tool selection drive strategy rather than the other way around. The correct sequence is to identify business weak points first, set KPIs that address those specific weak points, then build campaigns and select tools accordingly.

Tool selection itself should be governed by three factors: integration with the existing platform (including access to real-time data), compliance suitability for the markets the operator is active in, and any specific functional requirements that a given market or player segment demands. Operators on the Pragmatic Solutions PAM platform benefit from open API architecture that gives CRM teams genuine freedom of tool choice without sacrificing data quality or speed.

When generic campaigns outperform personalised ones

Personalisation is frequently treated as the default target state in CRM, but the conversation with Brigitta surfaces a more nuanced view. In the early stages of a player relationship, there is insufficient data to personalise meaningfully, and attempting to do so delays execution. A well-timed generic campaign will outperform a complex personalised setup that takes weeks to build and deploy.

The practical principle is that generic does not mean irrelevant. Broad segmentation, such as targeting all players of a given game type at a commercially appropriate moment, can be highly effective. Personalisation becomes valuable when it genuinely changes player decisions, not when it simply inserts a first name or a favourite game into a message template.

CRM and the rest of the business

A point Brigitta raises that often gets overlooked: CRM cannot compensate for problems that originate elsewhere in the product. If a player is dropping off because of a friction point in the payment journey, a UX issue, or a technical fault, increasing messaging volume may increase frustration.

Effective CRM therefore requires alignment with product and operations teams. When a campaign is underperforming, the first question should not be whether the campaign itself needs refinement. It should be whether there is an underlying product or operational issue that CRM activity is unable to address. This cross-functional accountability is often absent in operator structures where CRM sits in isolation.

VIP CRM as an early warning system

The standard framing of VIP CRM focuses on communication strategy: channels, cadence, personalised incentives. Brigitta reframes the function more usefully. CRM's primary role in VIP management is as a detection system. The signals that indicate a VIP player is at risk of churning are behavioural: sequences of heavy losing sessions, declining session length, slower response times, absence of previously consistent activity.

By the time a meaningful drop in deposits appears, the player may already have disengaged or moved to a competitor. Operators who monitor behavioural signals through CRM have a meaningful lead time to intervene. Those who wait for revenue metrics to flag the problem are responding too late. The Pragmatic Solutions platform provides the real-time player data that makes this kind of early detection operationally viable.

AI in CRM is not a replacement for strategy

The episode includes a direct discussion of AI's role in CRM operations. Brigitta's view is that AI is most effective in the roles that humans do poorly or slowly: detecting pattern changes in player behaviour, scaling campaign execution across channels, and eliminating the emotional bias that can lead a CRM manager to persist with a campaign that the data already shows is not working.

What AI has not replaced, at least not yet, is the strategic function: defining business objectives, diagnosing structural problems, and protecting brand integrity over time. AI also requires clean, reliable data to function well. In environments where data pipelines are delayed or inconsistent, AI-driven automation creates rather than resolves problems.

5 key takeaways

• CRM campaigns built around deposit conversion are measuring an intermediate step, not the outcome that drives long-term player value.

• The correct sequence is problem diagnosis, then KPI-setting, then campaign design. Building live campaigns and then checking whether they move revenue is backwards.

• A well-executed generic campaign deployed quickly will often outperform a personalised one that takes weeks to build. Speed has genuine commercial value.

• VIP churn is personal and statistically significant. The behavioural signals that precede it appear in CRM data before they appear in revenue figures.

• AI is a useful operational tool for pattern detection and bias reduction. It is not a substitute for strategic expertise or clean underlying data.

Watch the full episode here.