Cross-Sell Banner on Cashier Adds 2 Taps, 38% Abandon the Deposit

A cross-sell banner injected on the cashier screen added two taps to the deposit flow and preceded a 38% drop in completed deposits among affected sessions. The figure comes from a mid-2024 funnel audit covering roughly 41,000 deposit attempts across three Croatian-facing brands, and it held after controlling for traffic source and device. The mechanic is not exotic: a promotional tile placed between the amount selector and the confirmation step, with a dismiss control that was small, grey, and easy to miss.

What follows is a working breakdown of why a two-tap addition can cost more than a third of deposits, how to measure the damage properly, and what the Croatian market's specific conditions — payment mix, licensing constraints, player demographics — do to the numbers. If you run acquisition or product for a Croatia-licensed operator, the audit's findings are directly transferable; if you are a player, the mechanics explain some of the friction you have probably felt at the cashier and why it is worth knowing which operators treat the deposit screen as a checkout rather than a shop window.

The arithmetic of a two-tap tax

Two taps sounds trivial. On a desktop session it is perhaps 1.4 seconds. On mobile, where the same audit found the median deposit flow already ran to nine interactions before the banner was added, it is closer to 3.8 seconds plus a scroll correction on smaller viewports. The abandonment is not spread evenly across that extra time; it clusters.

The audit segmented the 38% figure by stage. Of the affected sessions:

• 11.2% dropped at the banner itself — the player tapped the tile expecting it to advance the deposit, landed on a promotion page, and did not return.
• 19.6% dropped after dismissing the banner, at the confirmation step. Time-on-step rose from a median 6.1 seconds to 9.4 seconds, and the extra dwell correlated with back-navigation.
• 7.4% dropped at the payment-method selection that the banner had pushed below the fold on 360px-wide screens.

That last segment is the one most operators underweight. On a 360×640 viewport — still a meaningful share of Croatian mobile traffic, particularly among players on older Android handsets — the banner consumed 118 pixels of vertical space. Card and bank-transfer options that had been visible without scrolling moved out of view. The audit's replayed sessions show players tapping where the payment method used to be, getting nothing, and leaving.

Why the second tap is worse than the first

The first tap is a decision point; the second is a recovery. Once a player has tapped a promotional tile by accident, the deposit is no longer a straight line but a small navigation problem, and navigation problems at the cashier are expensive because the player's intent is time-limited. The audit found that sessions where the player back-navigated at least once converted at 51.3% of the baseline rate, against 78.9% for sessions with a single mis-tap and no back-navigation. Two taps is the headline; the back-button is the actual damage.

What the Croatian payment mix does to the drop-off

Croatia's deposit mix is not the UK's or Malta's, and the composition changes how much a cashier interruption costs. Since the euro adoption in January 2023, the currency friction that used to absorb a few seconds of player patience is gone — deposits are denominated in euro, and players expect the flow to be as fast as any other euro checkout. The audit's baseline median deposit time was 34 seconds from cashier entry to confirmation; every added interaction is now measured against a much tighter expectation.

Payment method matters more than device. Roughly 62% of the audited deposits used cards, 21% used bank transfer or instant bank payment, and 9% used e-wallets; the remainder split across vouchers and other methods. The banner's placement penalty fell hardest on bank transfer, where the flow already involves a redirect and a return, and where the extra dwell at confirmation pushed the session past the bank's own timeout in 3.1% of cases. A player who returns from their banking app to find the cashier has reset does not usually try again the same evening.

There is a licensing dimension too. Croatia's regulator, Hrvatski zavod za kockanje (HZJK), requires operators to present responsible-gambling information at defined touchpoints, and the deposit flow is one of them. A cross-sell banner competing for the same screen space as the deposit-limit reminder is not just a conversion problem; it is a compliance risk if the reminder ends up below the fold or behind a dismissible overlay. The audit found three of the audited brand-versions placing the banner above the limit reminder on mobile. That is the kind of detail an HZJK inspection will find faster than your own QA will.

The self-exclusion and limit-setting overlap

Players with active deposit limits behave differently at the cashier, and the audit's most uncomfortable finding sits here. Among sessions from accounts with a daily deposit limit set, the banner's presence correlated with a 44.6% drop in completed deposits — higher than the 38% headline. The plausible reading is not that limited players are more impulsive; it is that a limited player arrives at the cashier with a specific amount in mind and a hard ceiling, and any friction that makes the flow feel less deterministic is more likely to end the session. The population is small — 4.2% of audited accounts — but it is the population you least want to frustrate, both commercially and from a safer-gambling standpoint. A player who abandons a deposit because the cashier felt confusing is a player who may not come back to set a limit next time.

How to measure the damage without fooling yourself

The 38% figure is only useful if you can reproduce it. Most operators cannot, because their cashier analytics stop at "deposit started" and "deposit completed" with nothing in between. The audit's method is worth copying:

1. Instrument every step, not every page. The cashier is usually a single-page app; page-view analytics will tell you nothing. Fire events on amount entry, banner impression, banner tap, banner dismiss, payment-method selection, confirmation, redirect, and return.
2. Segment by viewport height, not just width. The fold problem is vertical. A 360×640 device and a 414×896 device are both "mobile" but behave differently; the audit's drop-off concentrated in the shorter viewport.
3. Hold the banner for a clean control. The audit used a 50/50 holdout for eleven days across the three brands. Shorter windows were contaminated by payday cycles; the first four days of any month in Croatia show a deposit spike that swamps a small effect.
4. Count back-navigations as failures, not noise. A back-navigation at the cashier is a near-certain lost deposit in that session. The audit's 51.3% relative conversion rate for back-navigating sessions is the single most actionable number in the dataset.

A word on sample size: the 38% figure has a 95% confidence interval of roughly ±1.9 percentage points at the audit's volume. If your Croatian brand does 6,000 deposits a month, a two-week holdout will not resolve a 3-point effect. Either run longer or accept that you are making a directional decision, which for a banner is usually fine — the cost of removing a bad cross-sell is close to zero.

What the A/B test should actually compare

The obvious test — banner versus no banner — answers the wrong question. The commercial argument for cross-selling at the cashier is real: a well-placed promotion can lift sportsbook-to-casino or casino-to-sportsbook crossover by a measurable margin. The audit's own data showed a 4.1% lift in cross-product take-up among players who did engage with the banner, against a deposit loss that dwarfed it. The test worth running is placement: banner after confirmation versus banner before amount entry versus banner on the post-deposit landing screen. The last of these is where most of the industry has quietly moved, and the audit's post-hoc comparison of one brand that relocated its banner mid-test supports it — the relocated version recovered 71% of the lost deposits while retaining 63% of the cross-sell lift.

Designing the cashier so the cross-sell pays for itself

The lesson is not "never cross-sell at the cashier." It is that the deposit screen has a budget of attention and interactions, and a promotion spends from it. Three design rules follow from the audit data and from how Croatian players actually deposit.

Keep the confirmation step within one screen. The audit's baseline flow fit amount, method, and confirm on a single 360px-wide screen with no scrolling. That is the standard to defend. Any element that pushes confirmation below the fold — banner, trust badges, bonus terms link — should be treated as a conversion cost and justified against its measured lift.

Put the promotion after the money moves. The post-deposit screen is the highest-intent moment in the session and the lowest-risk placement. A player who has just deposited is engaged, has time, and — critically — is not mid-task. The audit's relocated-banner data suggests the cross-sell lift holds up here; the deposit loss does not occur because there is no deposit left to lose.

Make dismissal honest. A dismiss control that is small, low-contrast, or positioned as an afterthought is a dark pattern, and in Croatia it sits awkwardly against HZJK's expectations around clear commercial communication. The audit found that the 11.2% who dropped at the banner itself were disproportionately players who tapped the tile expecting it to advance the flow — a sign the tile looked like a button. If the banner must exist pre-confirmation, it should look like a banner: full-width, clearly labelled as an offer, with a dismiss control that meets the same contrast standard as the primary button.

The bonus-terms link nobody reads

One more placement detail from the audit deserves attention. On two of the three brands, the banner carried a "terms apply" link that opened the bonus conditions in a new tab. The audit's session replays show 6.8% of banner-tappers opening that tab and 2.3% of them never returning to the cashier tab — a small but entirely avoidable leak. Bonus terms belong on the promotion page, not on the deposit screen. If a player needs to read wagering requirements to understand an offer, the offer is being presented at the wrong moment.

Where this leaves the cashier

The 38% figure is a warning about a specific pattern — promotion inserted mid-task, on a screen with no vertical headroom, with a dismiss control designed to be overlooked — and it is a reminder that the cashier is the one screen in an iGaming product where the player's job is unambiguous and any deviation from it is read as an obstacle. The audit's own authors were careful to note that the effect is placement-dependent, not universal: the same banner on the post-deposit screen cost nothing measurable.

The open question is whether the industry's cross-sell ambitions can survive that relocation. Post-deposit placement converts well but reaches fewer players, because it only sees sessions that completed. Pre-confirmation placement reaches everyone and costs a third of them. The commercial calculation that produced the banner in the first place — cross-product take-up, lifetime value, the assumption that a deposit screen is free real estate — was made without the funnel data. Now that the data exists, the question for Croatian operators is whether the cross-sell team and the payments team are looking at the same dashboard. In most organisations, they are not, and the banner stays up because nobody owns the 38%.