Two-Step Deposit Prompts Cut Reversals 18% in 500 Wallets

A 500-wallet field trial across three licensed operators found that replacing the single-tap deposit confirmation with a two-step prompt—an explicit intent screen followed by a separate confirmation—reduced same-session deposit reversals by 18.4%. The effect held across card, bank transfer, and open banking rails, though it varied sharply by player tenure: the reduction was 24.1% among accounts younger than 30 days and 9.7% among accounts older than one year. The trial ran from 3 February to 14 April 2025 and covered 1.42 million deposit events, making it one of the larger controlled tests of interface friction published in the iGaming compliance literature to date.

The result matters because reversal behaviour sits at the intersection of two competing pressures. Regulators in several jurisdictions now treat the window between deposit and wager as a consumer-protection concern, while operators treat the same window as a conversion funnel that leaks revenue. A prompt that adds roughly 1.8 seconds to the deposit flow appears to reduce the leak without suppressing initial deposits—total deposit volume in the treatment cohort was statistically indistinguishable from control at 99.2% of baseline.

What the trial actually measured

The study defined a reversal as any deposit cancelled, voided, or returned to the funding source before the funds were staked, within a 24-hour window. This excludes chargebacks, which involve a different mechanism and a different timeline. The 18.4% figure therefore describes voluntary or semi-voluntary withdrawal of intent, not payment failure.

Wallets were randomised at the account level rather than the session level, which avoids the contamination problem that plagues session-level A/B tests in betting products. A player who sees the two-step prompt in one session and the one-tap flow in the next will carry expectations across sessions; account-level randomisation removes that path. The trade-off is that account-level assignment requires a larger sample to detect small effects, which is partly why the trial needed 500 wallets and 1.42 million events rather than the 50,000 events a session-level design might have used.

The prompt design

The treatment prompt had two screens. The first displayed the deposit amount, the payment method, and a single sentence stating that funds would be available for immediate wagering. The second screen required the player to confirm the amount again by tapping a button labelled with the exact figure—£50, not "Confirm." This second-screen design choice is worth isolating, because it differs from the more common pattern of a generic confirmation button.

A secondary arm of the trial tested a generic confirmation button against the figure-specific button. The figure-specific version produced a 21.3% reversal reduction versus 14.9% for the generic version, a difference significant at p < 0.01. The implication is that the friction is not merely the extra tap; it is the act of re-reading the number.

Why reversals fall, and why that is not automatically good

Three mechanisms plausibly explain the effect.

Mismatch correction. A portion of reversals are not deliberate cancellations but corrections of mis-keyed amounts. A player who intended £20 and typed £200 will reverse on seeing the figure restated. The trial data supports this: reversals in the treatment arm clustered disproportionately at round-number amounts above £100, the range where mis-keying is most costly.

Impulse interruption. The second screen inserts a pause of roughly 1.8 seconds. Behavioural research on cooling-off intervals suggests even short delays reduce impulsive commitment, though the effect sizes reported in laboratory settings are typically smaller than what this trial observed in a live product.

Reduced autopilot. Players who deposit frequently develop motor patterns that bypass deliberation. A two-step prompt interrupts the pattern. This mechanism predicts the tenure gradient the trial found: newer accounts, which have not yet formed the pattern, show a smaller relative effect from interruption—except the trial found the opposite, with newer accounts showing a larger reduction. That inconsistency is unresolved and is the most interesting open question in the dataset.

The compliance reading is straightforward: fewer reversals means fewer players depositing more than they intended. The commercial reading is more ambivalent. Operators earn nothing on a reversed deposit, but they also earn nothing on a deposit that never happens. The trial's finding that total deposit volume held at 99.2% of baseline is the load-bearing number here. If the two-step prompt had suppressed deposits by even 3%, the reversal reduction would be a net negative for the operator and would likely be abandoned regardless of its compliance appeal.

Where the effect breaks down

The 18.4% headline conceals substantial heterogeneity.

The open banking result is counterintuitive. Open banking deposits settle faster than card deposits, which should make reversal less feasible, not more resistant to the prompt. One explanation is that open banking users in the sample skewed toward higher-tenure accounts, which independently show smaller effects. The trial did not report a tenure-controlled breakdown by rail, which is a gap worth noting for anyone attempting to replicate the design.

The problem of measuring "intended" deposits

The trial assumes that reversal is undesirable. That assumption deserves scrutiny. A player who deposits £100, sees the restated figure, and reverses may be exercising exactly the judgement the prompt was designed to support. But a player who deposits £100, reverses, and then deposits £100 again thirty seconds later has simply been inconvenienced. The trial did not report re-deposit rates within the same session, so the proportion of reversals that were followed by an immediate re-deposit of the same amount is unknown. If that proportion is high, the 18.4% figure overstates the consumer-protection benefit.

This is not a fatal flaw. It is a measurement boundary that the next study should address directly, ideally by instrumenting the session timeline rather than the deposit event alone.

What a regulator should take from this

The trial's most defensible claim is narrow: a figure-specific confirmation screen reduces voluntary deposit reversals by a measurable margin without reducing deposit volume. That is enough to justify adoption as a default interface pattern in jurisdictions where reversal windows are treated as a consumer-protection feature rather than a payment-processing artefact.

It is not enough to justify a mandate. The tenure gradient—24.1% for new accounts against 9.7% for established ones—suggests the prompt does most of its work during the period when players are still calibrating their relationship with the product. A mandate applied uniformly would impose the 1.8-second cost on high-tenure players who gain little from it, and the trial offers no evidence on whether that cost accumulates into meaningful friction over hundreds of sessions.

The more productive question is whether the effect persists. Behavioural interventions of this kind frequently decay as users adapt: the second screen becomes as automatic as the first tap once it is familiar. The trial ran for ten weeks, which is long enough to detect early decay but not long enough to rule out a slower erosion. A twelve-month follow-up on the same 500 wallets, measuring reversal rates at months three, six, and twelve, would settle whether the 18.4% is a durable shift in behaviour or a transient response to novelty. Until that data exists, operators adopting the pattern should instrument their own reversal rates by tenure and rail, and treat the published figure as a prior rather than a target.