Cashier Back Buttons: 64% Return to the Withdrawal Screen Twice

64% of players who hit the back button on a cashier screen return to the withdrawal screen a second time within the same session — and a meaningful slice of them get stuck in a loop that ends with the pending withdrawal cancelled. That figure comes from session-replay analysis of roughly 410,000 cashier interactions across UK-licensed operators in the twelve months to March 2025, and it points at something operators rarely measure: the cost of a back button.

What the back button actually does

On most cashier builds, "back" isn't a neutral navigation control. It's a state change. Pressing it from the withdrawal confirmation screen typically drops the player to the deposit screen, sometimes with the deposit amount pre-filled and the payment method carried over. The player intended to check a detail — the fee, the processing time, the last four digits of the card — and lands somewhere that looks like an invitation to reverse.

The 64% figure is a return rate, not a reversal rate. Most of those second visits end with the withdrawal confirmed. But the sub-group that matters is the one that doesn't: players who bounce between withdrawal and deposit three or more times in a session cancel the pending withdrawal at roughly 2.7 times the baseline rate. That's the number a payments team should be looking at, not the headline return figure.

Why the loop forms

Three patterns account for most of the repeat traffic.

Fee and timing checks. A player who doesn't know the withdrawal will take 1–3 working days, or that a £2.50 fee applies below a threshold, will go back to look. If the fee is only shown on the deposit screen, the loop is a design fault, not player behaviour.

Verification friction. KYC prompts mid-withdrawal send players backwards to find out why. Where the reason is vague — "additional checks required" with no detail — the return rate climbs sharply.

Reversal prompts. Any element on the deposit screen that references the pending withdrawal, even obliquely, increases cancellation. The Gambling Commission has been explicit that reverse withdrawal functionality should be removed or made significantly harder to reach. A back button that lands on a pre-filled deposit form is a soft version of the same thing.

What the session data suggests

Operators that moved the withdrawal summary — amount, fee, expected arrival date, masked account details — onto the deposit screen saw return-to-withdrawal traffic fall by around 40% in the following quarter. That's a cheap fix: no new licensing work, no payment provider change, just putting the information where the player is looking for it.

The harder fix is the back button itself. Some operators now route "back" from the withdrawal confirmation to a read-only transaction summary rather than the deposit flow. Early numbers are thin, but the direction is consistent with what we already know about friction: remove the path of least resistance to reversal and cancellations drop.

Worth noting for anyone reading the 64% as evidence players are indecisive: they're not. They're checking. The question is whether your cashier gives them a way to check without walking back into the deposit flow.

The open question

If 64% of players go back to the withdrawal screen twice, what's the equivalent figure for players who abandon entirely and return the next day? Nobody in the sample was tracking cross-session return rates against cashier design, which means the true cost of a poorly built back button is probably higher than the 2.7x cancellation multiplier suggests. The next round of analysis should follow the session, not just the screen — and it should be published, because cashier UX is one of the few areas where operators could share findings without giving away anything commercially sensitive. Players who want to understand their own patterns can request their account history from any UK-licensed operator, and GamCare's helpline (0808 8020 133) is there if the loop stops feeling like a design problem and starts feeling like something else.