KYC Re-Checks at Each Withdrawal Lift Dropout 37% in 500 Files
A sample of 500 withdrawal files from three licensed operators shows that running a fresh know-your-customer check at every withdrawal, rather than once at onboarding, is associated with a 37% dropout rate before the payment is authorised. The figure covers a twelve-month window ending in March 2025 and compares against a 14% dropout rate on matched files where identity verification was completed at account opening and not repeated. The gap survives controls for withdrawal size, tenure, and payment method, though the study design cannot fully separate the friction of re-checking from the characteristics of players who trigger it.
How the sample was built
The 500 files were drawn from a pool of roughly 41,000 withdrawal requests processed by the three operators between April 2024 and March 2025. Selection was stratified by withdrawal band: 200 files between €50 and €250, 200 between €250 and €1,000, and 100 above €1,000. Within each band, half the files came from accounts subject to per-withdrawal verification and half from accounts verified once at onboarding, matched on account age (within 30 days), deposit history (within one quartile), and payment rail.
"Dropout" here means the player abandoned the withdrawal after submission and did not resubmit within 14 days. It excludes requests rejected outright for fraud or sanctions reasons, and it excludes players who completed the re-check within the session. That definition matters: a player who uploads a utility bill, waits nine minutes, and gets paid is not counted as a dropout, even though the friction may still cost the operator in future sessions.
The 37% figure is the abandonment rate on the re-check cohort across all three bands. The 14% figure is the matched baseline. The difference is 23 percentage points, or a relative increase of roughly 164% — the "37%" in the headline refers to the absolute rate on the re-check side, not to the size of the increase, a distinction worth keeping straight when the number gets quoted.
Where the dropout concentrates
Abandonment was not uniform. On files between €50 and €250, the re-check cohort dropped out at 41%. Between €250 and €1,000, the rate fell to 34%. Above €1,000, it was 29%. The pattern runs opposite to what a revenue-maximising operator might assume: smaller withdrawals are more likely to be abandoned, presumably because the expected value of persisting is lower relative to the effort of producing documents.
The baseline cohort showed a much flatter curve — 15%, 14%, and 12% across the same bands. In other words, the friction of re-verification does not simply scale with the withdrawal; it interacts with it, and the interaction is strongest where the amount at stake is smallest.
What the re-check actually involves
Across the three operators, "re-check at each withdrawal" meant different things, and the study treats them as one category only because the dropout rates clustered tightly.
- Operator A required a document upload (proof of address or source-of-funds evidence) on every withdrawal above €100, with manual review.
- Operator B ran an automated re-screen against identity and sanctions databases on every withdrawal, escalating to document upload only on a match or a threshold breach.
- Operator C required a fresh selfie-based liveness check on each withdrawal, plus document upload above €500.
Operator B, whose re-check was largely invisible to the player, had the lowest abandonment in the cohort: 22%. Operator C, with the liveness check, had the highest at 48%. Operator A sat at 39%. The spread suggests that the 37% aggregate is doing a lot of averaging, and that the mechanism — visible document demand versus silent re-screening — may matter more than the principle of re-checking itself.
That finding complicates any simple reading of the headline. If a regulator or compliance team takes "re-checking lifts dropout 37%" as an argument against re-checking, they may be drawing the wrong lesson. The data are more consistent with a narrower claim: visible, player-facing re-verification at the withdrawal stage is expensive in abandonment, while back-end re-screening is comparatively cheap.
The regulatory backdrop
Per-withdrawal verification is not arbitrary. The EU's Anti-Money Laundering Regulation, which applies from July 2027, tightens obligations around ongoing monitoring and, for certain risk profiles, verification at each transaction rather than at onboarding. Several national regulators — including those in the Netherlands, Denmark, and parts of Germany's interstate treaty regime — already require verification before a first withdrawal and, for higher-risk accounts, on an ongoing basis. Operators that read "ongoing monitoring" as "re-check every payout" are making a defensible compliance choice, but the data suggest they are paying for it in conversion.
The 500-file sample cannot adjudicate whether the compliance benefit justifies the cost. It was not designed to. What it can do is put a number on one side of the ledger, which is more than most operator compliance reviews do before implementing a workflow.
The confound worth naming
Players who trigger a re-check are not a random subset. They skew toward newer accounts, larger cumulative deposits relative to tenure, and payment methods with weaker built-in identity signals. Some of the 37% may therefore reflect the underlying risk profile of the cohort rather than the friction of the check. The matching controls reduce this but do not eliminate it. A cleaner design would randomise re-check timing within eligible accounts — ethically fraught, but not impossible under a regulator-supervised pilot, and the only way to isolate the causal effect.
What operators do with the number
The practical implication is not "stop re-checking." It is that the design of the check is a conversion variable, not just a compliance one. Three levers show up in the data:
Timing. Re-checks triggered at withdrawal are the most expensive point in the player journey. Operators that front-load verification — completing document collection at or shortly after first deposit, before the player has a withdrawal pending — report lower abandonment, though the sample here is too small to quantify the effect.
Invisibility. Operator B's automated re-screen cost 22% abandonment versus 48% for the liveness check. Where regulation permits silent re-screening, it appears to be the cheaper path. Where it does not, the gap is a reason to argue for it.
Thresholds. The concentration of dropout in the €50–€250 band suggests that applying re-checks to small withdrawals is the worst value trade in the workflow. A €75 withdrawal that triggers a document upload is unlikely to survive the friction, and the compliance benefit of verifying it is marginal relative to the €1,500 withdrawal that would clear the same check more readily.
The open question
The 37% figure will circulate, and it will be cited both by compliance teams defending re-checks as a necessary cost and by commercial teams arguing to relax them. Neither camp has the evidence it needs. What the 500 files establish is a correlation with a plausible mechanism and a wide spread across implementation styles. What they do not establish is whether the players who abandon are, on average, the ones the check was designed to catch — and if they are, the dropout is not a cost at all but the control working as intended. Until someone runs the randomised pilot, that question stays open, and the number stays a talking point rather than a finding.