What is Churn rate in iGaming

Churn rate is the percentage of customers a business loses over a defined period. It measures how well a product keeps the people it already has. In iGaming, churn usually counts previously active players who stop meeting an activity threshold, such as a real-money bet or a deposit, in the next period.

What is churn rate?

Churn rate, also called customer attrition rate, is the share of an existing customer base that stops using a product during a set period. The formula is the same across industries, but the definition of “lost” is not.

In a subscription business, a customer churns when they cancel. Online casino players rarely close their accounts. Instead, they stop logging in, depositing, or betting, and the account stays open. So the operator has to decide what counts as active.

Most casinos use one of three events: a qualifying real-money wager, a successful deposit, or a session. Pick one, document it, and apply it the same way every period. For example, a player who wagered in January and only logged in during February is active under a login rule and churned under a wagering rule.

The rule changes the result. Here, three rules apply to the same cohort of 2,000 players who wagered in January.

Nothing about the players changed. Only the definition did. Therefore, a report that says “churn is 10%” without naming the rule is incomplete.

Churn rate formula

Churn rate (%) = (Players lost during the period ÷ Players active at the start of the period) × 100

Players active at the start form the baseline cohort. For monthly wagering churn, that cohort includes everyone who placed at least one qualifying bet in the previous month.

Players lost are members of that same cohort who did not meet the activity rule in the current month. Count losses only from the starting population.

New players acquired during the period stay out of the denominator. They were not in the cohort, so they cannot churn from it. Otherwise, adding them makes churn look lower than it is.

The formula works for weekly, monthly, and quarterly windows. However, comparisons hold only when the activity definition and the window stay the same.

Close image

How to calculate churn rate: step-by-step example

An online casino measures monthly wagering churn. The figures below are illustrative, not industry benchmarks.

In January, 2,000 players placed at least one real-money bet. In February, 1,800 of them bet again, but 200 did not.

Churn rate = 200 ÷ 2,000 × 100 = 10%

The casino kept 1,800 of its January players, or 90%.

Next, add 500 players acquired in February. The active base grows to 2,300, but churn stays at 10%. The top-line number went up while the casino still lost one in ten existing players. So acquisition can hide that loss.

The calculation then takes five steps:

1. First, define the measurement period and the activity rule.
2. Then identify the starting cohort by unique player ID.
3. Count cohort members who did not meet the rule in the next period.
4. After that, divide churned players by the starting cohort.
5. Finally, multiply by 100.

Before the first report, decide how to treat duplicate accounts, closed or suspended accounts, and self-excluded players. Then keep those rules fixed across product, CRM, and finance teams.

Close image

Common mistakes that distort the number

Most calculation errors come from the same few places.

Churn rate vs retention rate

Churn and retention describe the same cohort from opposite sides. The former counts who left, while retention rate counts who stayed.

Retention rate = 100% − Churn rate

This holds only when both metrics use the same population, activity rule, and window. For example, Day-30 retention usually follows players acquired on one date, while monthly churn often compares all active players across two calendar months. Those two numbers will not add up to 100%, because they answer different questions.

A short example shows the gap. Say 1,000 players register in January and 280 of them place a bet on day 30. Day-30 retention is 28%. Monthly churn for the whole active base in the same month can still be 10%. Both numbers are correct. They describe different groups.

Types of churn in iGaming

A single churn figure rarely explains much, so operators split it by the activity or the money they track.

Player, deposit, and wagering churn

Player churn is also known as logo churn: the share of individual customers lost, regardless of how much they spent.

Deposit and wagering churn can move apart. For example, a player can keep betting from an existing balance without depositing. Another can deposit and place no bet in the window. Neither metric substitutes for the other.

Revenue churn

Revenue churn weights players by value. A casino can lose 15% of its active players with little change in net gaming revenue (NGR) if those players were low-value. Losing a handful of VIPs does the opposite. Revenue churn needs a fixed cohort, a stated revenue base (GGR or NGR), and consistent accounting. In casino, payout variance moves cohort NGR month to month, so one weak month does not prove attrition.

Two formulas separate lost customers from changes in spend among customers who stay. Contraction is the drop in revenue from players who stayed active, while expansion is the increase.

Close image

Take the 2,000 January players again. They generated $400,000 in NGR. The 200 players who left generated $20,000 of it, so gross revenue churn is 5%, half the player churn rate. Meanwhile, players who stayed reduced their NGR by $15,000 in total and increased it by $45,000.

Net revenue churn = (20,000 + 15,000 − 45,000) ÷ 400,000 × 100 = –2.5%

The result is negative, because revenue from the starting cohort grew even though 10% of players left. Illustrative example.

Cohort churn

Cohort churn groups players by registration month, first-time deposit (FTD) date, channel, or market. Cohort analysis can show that two affiliate campaigns with similar sign-up volumes produce very different 30-day wagering churn.

Registration counts look identical, but the churn gap points to a difference in player quality or expectations that totals hide.

One boundary matters here. A registered player who never makes a second deposit is a conversion problem, not churn. Instead, track second-deposit conversion as its own metric.

What is a good churn rate?

No universal churn benchmark exists for online casinos. Published figures rarely state the activity rule, the window, or the sample, so they cannot be compared with an operator’s own number.

The same 10% means different things depending on:

• Measurement period. A player who skips a week may still play within the month, so weekly and monthly churn are not comparable.
• Activity definition. Wagering, deposit, and login rules produce different numbers from the same player base.
• Product. Sportsbook activity follows the sports calendar, while casino play follows its own patterns.
• Acquisition source and cohort maturity. New players behave differently from players with six months of history.
• Seasonality and market. Regulation, payment options, and competitor activity shift participation.

How published sources define churn

Published sources show how far the definitions drift.

The cut-off alone ranges from 30 days to six months. As a result, a rate calculated under one definition says little about a rate calculated under another.

Why an internal baseline works better

The useful benchmark is internal: one fixed definition, equivalent cohorts, and comparisons with the same period in previous years.

Churn also needs commercial context. Dmitry Belianin, Co-founder of Blask, described reviewing a site that was proud of its gross gaming revenue (GGR). However, the deeper analysis told a different story: “Churn was high. Player value was unpredictable. Retention was inconsistent.” GGR can grow while the player base weakens, which is why churn belongs next to ARPU, ARPPU, and player lifetime value (LTV).

Why do online casino players churn?

A rising churn rate shows that behavior changed, but it does not show why. Each factor below is a hypothesis, paired with the data that can test it.

Payments deserve a closer look because they control access to the player’s money. Compare players who hit a failed deposit or a slow withdrawal with similar players who did not. A gap in later activity justifies an investigation, but it does not prove that payment friction caused every exit.

That distinction applies to every row in the table. A factor becomes a demonstrated cause only when a controlled test or a matched comparison isolates it.

Some inactivity is the intended outcome, since players who self-exclude, set limits, or decide to stop gambling are not a retention target. The UK Gambling Commission requires operators to remove self-excluded customers from marketing databases within two days of notification, including databases used by third parties. Keep these players out of reactivation campaigns.

When pre-churn signals are harm signals

Some behavior that precedes inactivity is also a risk signal. The UK Gambling Commission’s formal guidance for remote licensees, in effect since October 31, 2023, is direct about marketing to these customers.

In addition, the same guidance lists the signals licensees should monitor: customer spend, patterns of spend, time spent gambling, gambling behavior indicators, customer-led contact, use of gambling management tools, and account indicators.

Some of these behaviors can also show up in churn models. Check any “at-risk” list against harm indicators before it reaches the CRM team.

How to reduce churn rate in iGaming

Start with the cause, not the bonus budget. Product fixes remove friction for every player. Promotional incentives change behavior while they run and add cost. Measure the two separately.

• Improve onboarding. Map every step from registration to the first qualifying bet. Track verification completion, deposit success, and FTD conversion. The outcome to watch is the share of new players who reach first activity.
• Optimize casino navigation. Review lobby discovery, categories, search, and game placement. Measure launches from the lobby, search success, and launch failures.
• Segment players. Compare churn by cohort, acquisition channel, product category, GEO, and device. A rise concentrated in one segment is easier to diagnose than a site-wide average.
• Improve CRM relevance. Test communications on eligible players with a holdout group. Opens and bonus redemptions do not prove retention, but incremental activity against the holdout does.
• Reduce payment friction. Monitor failed transactions by payment method and withdrawal times. Fix the methods with the highest failure rates first.
• Evaluate with controls. Use A/B tests or matched cohorts. Churn moves with the season, so a simple before-and-after comparison can credit a change with a result it did not cause.

Why timing matters: reactivation data

Optimove analyzed 5,341,332 players from operators worldwide between October 2023 and October 2024. It defines a churned player as one with 30 or more days of inactivity after deposit activity. Its research reports that 27% of churned players can be reactivated on day 1 after churning, against 2% after three months. Those who return after three months have 87% lower predicted future value.

Among players who did return, 60% came back with a new deposit and 40% with existing or bonus money. In addition, Optimove estimates that players who return by depositing have 44% higher future value than those who return without a deposit.

Because this is vendor research on proprietary data with Optimove’s own 30-day definition, treat it as a pattern to test in your own cohorts, not as a benchmark. Speed applies only to eligible players, and it does not apply to anyone showing harm indicators or who has self-excluded.

Close image

Cases: what operators and vendors report

YesPlay (South Africa). In an Optimove case study, the operator reports that a year after adopting the platform its churn fell 17%, unique monthly players rose 334%, net revenue rose 354%, and total bet amount rose 8%. The source does not define churn, and the operator was growing fast, so the churn change cannot be credited to the platform alone.

MrQ, a UK online casino, described the problem behind its retention work in a Dotdigital case study published in June 2021.

Operator interviews also show a similar gap. Ricardo Rosada, founder of BR Marketing, told Blask’s Signal and Noise podcast that many Brazilian operators underuse the data they already hold, as published on the Blask blog.

What research says about prediction

Academic work on casino churn prediction is limited but real. Suh and Alhaery applied a decision tree to online casino data and showed how classification rules and propensity scores can flag players at risk. Merchie and Ernst trained neural networks on Gaming1 data and found that models stay accurate on later dates. Their results still varied because of one-off events in the data.

A model flags who may leave, but it does not say why, and it does not test whether an intervention works.

A True Odds interview with Ilya Batcherikov, CPO at Blask, makes a related point: most products measure conversion and churn but miss everything in between. For an operator, the in-between is onboarding completion, payment errors, game discovery, and support contacts. Those signals change before a player goes inactive.

Close image

How to analyze churn alongside market demand

Churn rate comes from an operator’s own player data, but some of its drivers sit outside the platform: market demand, the sports calendar, regulation, and competitors. External market intelligence adds that context.

Blask Index measures iGaming demand for brands and markets from processed search signals. It does not see individual players and does not measure any operator’s churn. Instead, it shows the market conditions those players were in.

Blask also reports Brand’s Accumulated Power (BAP), a brand’s share of accumulated market demand. Acquisition Power Score (APS) estimates new customers within its methodology, and Competitive Earning Baseline (CEB) is a modeled revenue benchmark. However, none of them replaces internal retention, churn, or revenue data.

Case 1: Brazil after the September 2026 ban

Brazil in autumn 2026 shows why the outside view matters. On September 25, 2026, Brazil banned its licensed betting market. Before that, Brazil’s Blask Index averaged 7.22M a day from September 2 to 24. Then, in the week of October 19–25, it averaged 292K a day, down 96%.

Any licensed operator in Brazil saw churn spike in October. The cause was regulatory, so no onboarding fix or CRM campaign would have changed that number.

Close image

Case 2: Germany, licensed and offshore demand

Germany shows a quieter version of the same problem. Blask Index for licensed brands held steady between September 2025 and September 2026, while demand for offshore brands more than tripled.

Close image

A licensed operator with flat absolute demand can still see its active base thin out when its share of market attention falls. The data cannot show where individual players went or why, because Blask Index measures search demand for brands, not player activity.

One regulatory factor is worth testing in deposit-churn cohorts. Germany’s State Treaty of 2021 caps deposits at €1,000 per calendar month across all licensed operators, according to the regulator GGL. GGL says higher individual limits are possible partly to stop players moving to illegal offerings. Still, whether the cap contributes to inactivity at any given operator is a hypothesis, not a finding.

Two patterns to test

Most cases are less extreme than Brazil, so two patterns help frame an investigation:

• Churn rises while market demand falls. External demand becomes one candidate explanation. Internal data still has to confirm it.
• Churn rises while market demand holds. The operator is losing its own players in a stable market. Product, payments, acquisition quality, and competitors come first.

Churn rate analysis: practical operator workflow

Illustrative scenario. Monthly wagering churn rises from 10% to 13%. That is 3 percentage points, or a 30% relative increase. The analytics team then works through six steps.

1. Verify the methodology. Confirm that the activity rule, windows, eligibility rules, and data pipeline did not change. A tracking error can produce a spike that never happened.
2. Identify affected cohorts. Split churn by acquisition month, channel, GEO, product, and device. Establish whether the rise is broad or concentrated.
3. Compare product, payment, and activity data. Check failed deposits, withdrawal times, game launch errors, and session trends for the affected cohorts.
4. Investigate operational changes. List releases, lobby updates, provider outages, bonus policy changes, and payment provider switches in the period.
5. Review external market conditions. Compare the timing with market demand in Blask Index, the sports calendar, and regulatory news. External signals suggest explanations, but they do not confirm them on their own.
6. Test interventions and measure the effect. Say a navigation update coincided with more failed game searches in the affected cohort. Fix it for part of the audience, keep a control group, and compare churn on the same definition. Track responsible gambling indicators alongside the result.

Step 2 often shows that the rise is not broad. In this illustrative case, one channel accounts for the whole increase.

The first hypotheses now concern one channel: its onboarding path, the payment methods its players use, and any change in the offer it promotes.

Close image

Frequently asked questions

Definition and calculation

Benchmarks, negative churn, and reduction

Inactivity cutoffs and revenue churn

Conclusion

Churn rate tells an operator how much of its active player base it lost. The number is only as reliable as its definition, so the activity rule, the window, and the starting cohort have to stay fixed.

The percentage shows that something changed, but finding out what changed takes cohort splits, product and payment data, and a view of the market outside the platform. Internal player data explains who left, while market demand data shows the conditions they left in.

See the market demand behind your player dataTry Blask