Series: Rethinking CX in iGaming — Part 2 of 4
By: Anamarija Perović, Head of Customer Experience @ Ludicrum, in collaboration with Cevro AI

Imagine your customer support team closes out another strong month. CSAT sitting at 4.7 out of 5. Response times down. Agent productivity up. Every metric in the dashboard is pointing green. And then you look at your churn report.

This is the paradox more iGaming operators are quietly wrestling with. The explanation almost always comes down to which questions you're asking — and whether the metrics you're trusting are actually equipped to answer them. CSAT and Voice of the Customer don't measure the same thing, and using one as a proxy for the other is one of the most common — and most costly — blind spots in iGaming CX today.

High CSAT can create a dangerous illusion: that the experience is healthy, when in reality support teams have simply become very good at absorbing friction created elsewhere in the product.

CSAT Is a Lagging Comfort Metric

CSAT measures satisfaction with an interaction. Not with the product. Not with the journey that led a player to contact support in the first place. The interaction, that specific moment of contact between a player and an agent.

By the time a CSAT score is generated, the friction has already occurred. A player encountered a problem, decided it was worth raising, waited for a response, and — if you're lucky — filled in a survey. CSAT scores the recovery. It tells you how well you apologised. It doesn't tell you why you had to.

That's not a reason to abandon it. CSAT remains essential for managing agent performance, QA benchmarking, and day-to-day operational health. But it is a reason to be precise about what it cannot see — and to understand that the players creating your churn problem are, structurally, invisible to it.

The Hidden Cost of a High Score

Here's a scenario that plays out in iGaming support operations every day.

A contact category sits consistently in your top five by volume — something like "Why am I not eligible for this promotion?" Agents handle it well. CSAT on this category is strong. Leadership sees the numbers and moves on.

But consider what that contact category actually represents. Every ticket carries an agent cost, a tool cost, a supervision cost, a QA cost, an opportunity cost — because that agent could be handling a complex withdrawal dispute or a responsible gambling concern — and an emotional cost, because repeating the same explanation dozens of times a day is a quiet but real driver of agent burnout.

Now ask the question CSAT never prompted: why does this category exist at scale? Maybe the promotional communication is unclear. Maybe the bonus eligibility logic is confusing. Maybe the UX buries the information players need to self-serve. No one investigated it, because the score looked fine.

CSAT optimises the cost of handling a ticket. VoC eliminates the reason that ticket exists. The ROI difference between those two outcomes is not marginal.

The Silent Majority Problem

There is a structural flaw in CSAT that rarely gets discussed openly: it only measures the players who bother to respond — and with typical survey response rates sitting at 10% or lower, that is a vanishingly small and deeply unrepresentative sample.

The scale of what sits beneath that surface is striking. For every customer who complains, 26 others remain silent (HBR) — they don't fill in your survey, they don't reply to your follow-up email, they simply stop depositing and move on. According to PwC, 32% of customers will stop doing business with a brand they loved after just one bad experience. After several, that figure climbs to 59%. These players don't generate tickets. They don't generate CSAT scores. They generate churn.

This is the population CSAT structurally cannot see. A score built on a 10% response rate, skewed toward the vocal minority, tells you almost nothing about the players who experienced friction, said nothing, and quietly left. In iGaming — where player acquisition costs are high and lifetime value is everything — the cost of misreading this signal is not marginal.

Where VoC Is Different

This is exactly the gap VoC is designed to close, and where the shift in framing changes everything.

VoC doesn't ask "how did we handle this interaction?" — that's CSAT's question. VoC asks "why does this interaction exist, and what would it take to eliminate it?". Instead of measuring whether agents resolved tickets efficiently, it asks whether the product, the UX, and the communications are generating unnecessary contacts in the first place. Support stops being an operational cost centre and becomes a strategic signal for the rest of the business.

The commercial impact of that shift is well documented. Companies that lead in customer experience outperform competitors by nearly 80% in revenue growth, according to the Qualtrics XM Institute. Those are not gains driven by faster handle times or higher CSAT scores — they are structural gains, driven by understanding and removing the friction that causes players to disengage in the first place. VoC is the mechanism that makes that understanding possible.

VoC at Scale in AI Support

The reason VoC has historically been underutilised in iGaming is not a lack of awareness — most experienced CX leaders understand its value. The barrier has been execution. Traditional VoC programmes are sample-based, slow, and resource-intensive. By the time the analysis is complete, the patterns identified may already have shifted.

AI changes this entirely. An AI customer support agent operating at full deployment processes every interaction — not a sample. It tags intent, clusters contact reasons, and surfaces patterns in real time rather than in quarterly reports. VoC stops being a periodic exercise and becomes a live operational signal.

Returning to the bonus eligibility example: an AI system processing contact data at scale would flag the growing volume of bonus-related queries, link them to a specific promotion or eligibility rule, and surface that pattern to product, marketing, and CRM teams with enough context to act before it becomes a material cost or a churn signal. The support data stops being a record of what went wrong and starts being an early warning system for what needs to change.

Picking the Right Metric for the Right Question

CSAT and VoC are not in competition. They answer different questions, and the operators building genuinely efficient, player-centric support functions use both — deliberately, and for different purposes.

CSAT answers: did we handle this well? It belongs in your daily operational toolkit, your agent performance frameworks, your QA processes. VoC answers: should this have happened at all? It belongs in your product roadmap, your UX reviews, your CRM and promotional strategy.

The most expensive support interaction is the one that happens repeatedly, and goes unchallenged because the CSAT score looks fine.

In Part 2 of this series, we examine why CSAT scores give you the score but rarely the reason — and why Voice of Customer data is where the real ROI lives.

Anamarija Perović is a seasoned Customer Experience professional with deep roots in the iGaming industry. Currently Head of Customer Experience at Ludicrum.