Payment Icons Reorder at kn 0.10, 37% Tap the Wrong Rail

A 0.10 kn reorder of payment icons in a Croatian-facing cashier changed which deposit rail players actually tap, and the shift was not in the direction the layout team predicted: 37% of users in the test cell selected the rail that had moved, not the one that had been promoted to the top-left slot. The finding comes from a checkout-usability audit run across a mid-size operator's Croatian traffic in Q1 2025, covering 41,600 deposit sessions, and it matters because payment icon order is one of the few cashier variables that costs nothing to change and is almost never measured properly.

What the 0.10 kn Test Actually Measured

The test was not a bonus test, a bonus-amount test, or a wagering test. It was a pure layout intervention: the operator moved its three highest-volume deposit rails — card, bank transfer, and a local instant-banking option — into a new left-to-right order in the cashier's payment-icon strip, and priced the change at a nominal 0.10 kn per session in internal accounting terms to isolate the cost of the UI change from any promotional spend. No bonus was attached. No odds were altered. The only variable was which icon sat first.

That last detail is the whole point. Croatian cashiers have spent years optimising the things players are told to care about — welcome bonus size, free-spin counts, minimum deposit thresholds — while the icon strip that greets every depositor before any of that matters has been treated as a design afterthought. The audit suggests the opposite priority is warranted. When the operator rotated its icon order, 37% of sessions in the affected cell ended on a rail that was not the operator's preferred default. In the control cell, where the old order held, that figure sat at 22%.

Why "wrong rail" is not the same as "bad outcome"

The 37% figure needs a caveat before anyone treats it as a failure metric. Tapping a non-default rail is not inherently bad — if a player's bank card is the only instrument they have loaded, sending them to a bank-transfer screen wastes their time and yours. The audit's own definition of a "wrong rail" selection was narrower: a tap on a payment method the player had never previously used, or one that resulted in an abandoned deposit within 90 seconds. By that definition, 37% of the reordered cell's sessions involved a mis-tap or a bail-out, against 22% in control. The gap — 15 percentage points — is the number worth arguing about.

The Croatian Payment Stack Is Unusually Fragile to Layout

Croatia's deposit mix differs from the Western European default in ways that make icon order more consequential than it would be in, say, the UK or Germany. Card penetration is high, but the local instant-banking rails carry a disproportionate share of smaller, faster deposits, and players move between them based on habit rather than fee comparison. A Croatian player who deposits 200 kn on a Friday night is often not choosing a rail in any deliberative sense — they are tapping the icon their thumb already knows.

That muscle memory is exactly what a reorder disrupts. In the audit's session recordings, the pattern was consistent: players who had used the cashier before the change frequently tapped the position where their usual rail used to sit, landing on whatever icon had moved into that slot. The icon itself had not changed. Its label had not changed. Only its coordinates had. And that was enough to produce a 15-point swing in mis-selection.

The 41,600-session sample, and what it does not prove

The sample is large enough to be interesting and small enough to be cautious about. 41,600 deposit sessions across one operator's Croatian traffic is a real dataset, not a blog anecdote, but it is a single operator, a single quarter, and a single reorder. It does not establish that icon order matters equally across all Croatian-facing brands, nor that the 37% figure generalises. What it does establish is that the effect was large enough to detect without a controlled experiment, which is a low bar that most cashier redesigns still fail to clear.

The audit also did not isolate device type. A meaningful share of Croatian deposit sessions happen on mobile, where the icon strip may render as a horizontally scrollable row rather than a fixed grid — a layout difference that could amplify or suppress the position effect. The operator's own logs flagged mobile sessions as overrepresented in the mis-tap group but did not publish a device-split breakdown.

Where the Real Cost Sits

The obvious reading of the audit is "don't reorder your payment icons." That reading is wrong, and it misses where the money actually goes.

Abandoned deposits versus support tickets

An abandoned deposit is the visible cost. A player taps the wrong rail, sees a screen they did not expect, and closes the cashier. The operator loses the deposit, and depending on the rail, may also lose the player to a competitor they open in the next tab. The audit's 90-second abandonment window is a reasonable proxy for this, but it almost certainly undercounts slower bail-outs — players who navigate back, get confused, and leave two minutes later.

The invisible cost is support load. Every mis-tap that reaches a payment screen the player cannot complete generates a fraction of a support contact: a live-chat query, an email, a callback request. Croatian-facing operators typically staff Croatian-language support during peak hours, and deposit-related contacts are among the most common ticket types. A 15-point increase in mis-selection does not translate linearly into a 15-point increase in tickets — most players self-correct — but even a 3-4% conversion from mis-tap to contact is material at scale.

The reorder was probably still correct

Here is the uncomfortable part. The operator that ran this audit almost certainly had a good reason to reorder its icons: pushing a cheaper rail, a rail with better approval rates, or a rail with faster settlement. Those are legitimate commercial objectives, and a 15-point mis-selection rate does not automatically make the reorder a mistake. The question is whether the operator measured the trade-off.

Most do not. Icon order gets changed as part of a broader cashier refresh, bundled with branding updates or a new payment provider integration, and the resulting deposit-mix shift is attributed to the provider rather than the layout. The audit's contribution is not "reordering is bad" — it is "reordering has a measurable effect, and you should measure it."

What a Sensible Test Looks Like

If icon order is going to be treated as a real variable, it should be tested like one. That means a few unglamorous things.

Hold the order stable for a control group. The audit's control cell existed because the operator happened to run a staggered rollout. That was luck, not design. A deliberate holdback — even 10% of traffic — turns a redesign into an experiment at almost no cost.

Define the failure metric before the test. "Wrong rail" needs a definition that survives contact with the data. The audit's version — unused rail or 90-second abandonment — is defensible but crude. A tighter version would track whether the player completed a deposit within the session, on any rail, and whether they returned within seven days.

Segment by device and by player tenure. A first-time depositor and a five-year regular respond to icon position differently. The regular has muscle memory; the new player is reading labels. Aggregating them into one 37% figure hides the effect that matters most.

Do not change anything else in the same window. This is the hardest one. Cashier redesigns rarely happen in isolation. If the icon reorder ships alongside a new bonus, a new provider, or a new KYC step, the resulting data is uninterpretable.

The regulatory backdrop is not the constraint here

Croatian operators work under a licensing and reporting regime that governs what they can offer and how they advertise it, not how they arrange buttons on a deposit screen. There is no rule requiring a particular icon order, and no supervisory expectation that the cheapest rail sits first. That means the layout question is entirely a commercial and player-experience one — which is precisely why it should be tested rather than assumed. It also means the responsible-gambling angle sits elsewhere: deposit-limit tools, cool-off periods, and self-exclusion flows need to be reachable, and burying them behind a cluttered payment strip is a worse design failure than putting the wrong rail in slot one.

The Question the Audit Leaves Open

The 37% figure will get quoted as evidence that reordering payment icons is dangerous. That is the wrong lesson. The right lesson is that a 0.10 kn layout change moved player behaviour by 15 percentage points, and nobody predicted it — which means the cashier is full of similar unmeasured variables. Icon order is one. Icon size, label wording, the position of the "other methods" expander, the colour of the selected state, and the number of rails shown before a scroll are all candidates for the same treatment.

The open question is whether Croatian-facing operators will start treating the cashier as a testable surface or continue treating it as a branding exercise. The audit suggests the former pays. It does not yet prove by how much, and the operator that ran it has not published the deposit-completion numbers that would settle the argument. Until someone does, the 37% stands as a warning about a variable most teams still cannot see — and a reminder that the cheapest change in the cashier is often the one nobody bothers to measure.