Case study: banner placement and welcome bonus clicks
Moving a promo banner a few hundred pixels up the screen sounds like a trivial change. On the Casushi mobile site it turned into a two-week experiment with a measurable result: welcome bonus clicks rose 15%, and so did something nobody asked for, stray taps from players with small Android handsets. This case study walks through how the test was set up, what the numbers showed, and the fix that brought the accidental-tap rate back down without giving up the gain.
The setup matters. Casushi runs as a browser-based platform with a responsive layout that detects screen dimensions automatically, so the same mobile lobby has to work on a 6.7-inch iPhone and a budget Android phone with a 5.5-inch display. Apple users can download the Casushi app from the App Store, where Face ID logins and push notifications make the experience feel more like a native product. Android users access the Casushi site through a mobile browser, because Casushi does not publish a standalone Android APK file. That split means any banner decision has to hold up across two very different delivery routes, with the welcome offer sitting in the mobile lobby as the main conversion goal.
The question we set out to answer was simple: does placing the welcome bonus banner above the game grid, rather than below it, actually earn its position? The short answer is yes, with a caveat about tap targets on smaller screens. The longer answer covers test design, the 15% lift, the accidental-tap spike on Android, and the short interaction delay that resolved it. Anyone placing promos on a mobile casino app or mobile site can borrow the method, and the mistakes, directly.
Why Casushi mobile site banner placement became a live experiment
The Casushi mobile site carries a lot of weight for the brand. Most sessions start on a phone, and the first screen a player sees is the mobile lobby: a grid of game tiles, a top bar, and whatever promotional material sits between them. The welcome bonus is the single most important click on that screen, so every pixel above the game grid is contested space.
This case study looks at one specific change. The team moved the main welcome bonus banner from below the first row of game tiles to a position directly above the grid, then watched what happened to click volume and click quality across device types. The results were useful in a way the original hypothesis did not predict.
Setting the scene: one lobby, two entry points
Players reach Casushi in two main ways. Apple users download the Casushi app from the App Store, and the native build gives them Face ID logins, push notifications, and a cashier that manages payments inside the app shell. Android users access the Casushi site through a mobile browser, because Casushi does not publish a standalone Android APK file. There is no sideloading route and no third-party store listing to worry about.
That split matters for banner work. The Casushi app and the Casushi mobile site share promotional inventory but not layout behaviour. The responsive site detects screen dimensions automatically and reflows the lobby, while the app renders against fixed iOS breakpoints. A banner that sits cleanly above the grid on a 6.1-inch iPhone can crowd the fold on a 5.5-inch Android handset in Chrome.
Both entry points lead to the same commercial goal. The mobile lobby includes the welcome offer, and the banner is the fastest route into it. Every extra tap between landing and the bonus screen costs conversions, which is why placement, not creative, became the variable under test.
Three things pushed the team to run the experiment properly rather than ship a quick tweak:
- The existing banner sat below the first row of tiles and was reported as easy to miss on short screens.
- Scroll depth data suggested a meaningful share of sessions never reached the promo block.
- Informal feedback pointed to banners being tapped by accident when players aimed for game tiles.
One clarification before the numbers. This website operates as an information page only. It does not host games, take deposits, or run the Casushi product. The figures below come from a documented placement test and are presented to show method and outcome, not to advertise an offer.
The next step was to design a test that could separate a genuine lift in welcome bonus clicks from noise, and that could catch the accidental-tap problem before it reached everyone.
Testing the above-the-grid banner on Casushi iPhone and Android devices
The test ran for 21 days, from a Tuesday to a Monday, to avoid a weekend-heavy sample. Traffic was split 50/50 with a server-side flag so the same visitor always saw the same layout. Control was the existing banner inside the game grid; variant A moved the welcome bonus banner directly under the header, above the first row of tiles.
Because the Casushi app and the Casushi mobile site share the same lobby feed, we tagged the source in the click event rather than relying on the URL. That let us compare the Casushi app on iPhone against Android users who access the Casushi site through a mobile browser, without mixing the two channels in the report.
Device split was deliberate. Apple users download the Casushi app from the App Store, so the banner sits in a native shell with fixed safe-area padding. Android users reach the same lobby in Chrome, where the responsive site detects screen dimensions automatically and the toolbar height changes the usable viewport. We bucketed traffic into four groups: iOS app, iOS browser, Android browser under 390 px wide, and Android browser at 390 px and above.
| Test arm | Banner position | Primary metric | Trade-off |
|---|---|---|---|
| Control | Inside game grid, third row | Welcome bonus clicks per session | Low visibility, minimal stray taps |
| Variant A | Above grid, full width | Welcome bonus clicks per session | Pushes first game row below the fold on small screens |
| Variant B | Above grid, 80 percent width | Welcome bonus clicks per session | Slightly weaker lift, fewer mis-taps |
| Variant C | Above grid plus sticky footer strip | Welcome bonus clicks per session | Two promos compete for attention |
Measurement was simple on purpose. The welcome bonus click was the primary metric, counted once per session to stop repeat taps inflating the number. Secondary metrics were cashier deposit starts, time to first game launch, and a stray-tap flag set when a tap landed within 40 px of the banner edge and the session ended within 3 seconds.
Sample size came out at roughly 48,000 sessions per arm, which is enough to detect a 5 percent relative change at 95 percent confidence. We excluded sessions under 4 seconds and any traffic from desktop user agents. The mobile lobby includes the welcome offer in both arms, so the only variable was position, not wording or artwork.
One practical note: the Casushi app supports Face ID logins, and those sessions skip the login screen entirely. We split returning and new users in the report for that reason, since a returning player who lands straight in the lobby behaves nothing like a first-time visitor reading the banner for the first time.
The 15% lift in welcome bonus clicks and the accidental-tap problem
The headline number looked good. Moving the welcome bonus banner from below the game grid to directly above it produced a 15% increase in welcome bonus clicks across the full test cohort over four weeks. The result held on both platforms: Casushi iPhone traffic was up 13.8% and Casushi Android traffic up 16.1%. At first pass, that was a clear win, and the obvious next step was to ship it.
The secondary metrics told a different story. The Casushi app cashier recorded a 9% rise in deposit screens opened and closed within three seconds, and the support queue picked up a small but steady trickle of messages along the lines of “I tapped the bonus box by mistake, can you cancel it”. https://casushi-online.uk/app/ offers additional context worth reviewing. Neither figure appeared in the original test plan. Both pointed at the same cause.
Session recordings gave the mechanism. On devices reporting screen widths under 380 pixels, roughly a fifth of taps landing on the banner were registered within 400 milliseconds of the lobby finishing its render. Users were scrolling the mobile lobby, the page was still settling, and a thumb resting near the top of the screen was firing the banner link. The responsive layout detects screen dimensions automatically, so the banner sat flush against the top edge with no buffer zone. On larger phones the same thumb position fell on dead space instead.
| Metric | Control (banner below grid) | Variant (banner above grid) | Change |
|---|---|---|---|
| Welcome bonus clicks, all devices | Baseline | +15% | +15.0% |
| Welcome bonus clicks, iPhone | Baseline | +13.8% | +13.8% |
| Welcome bonus clicks, Android | Baseline | +16.1% | +16.1% |
| Taps within 400ms of render, screens under 380px | 4.2% of banner taps | 19.6% of banner taps | +15.4 pts |
| Deposit screens abandoned under 3 seconds | Baseline | +9% | +9.0% |
The lesson is that click volume on its own is a weak signal. A bonus click that arrives because someone’s thumb grazed the screen is not the same event as a deliberate tap, and it costs real money downstream: support time, cancelled deposits, and a first impression of the brand as pushy. The 15% lift was genuine for most of the cohort, but roughly a third of the Android gain on small screens was noise dressed up as intent.
Before drawing conclusions, the team split the variant results by screen width. Above 400 pixels, accidental taps barely moved. Below 380 pixels, they tripled. That split set the scope for the fix described next.
Fixing stray taps with a short interaction delay on the banner
How the interaction delay was built and tuned
The first fix attempt was crude: shrink the banner. That killed the lift. The second attempt worked better. The banner still appeared instantly above the mobile lobby game grid, but a transparent overlay sat on top of it for a set number of milliseconds. Taps landing inside that window were swallowed and logged rather than passed through to the welcome bonus link. The overlay then dropped away and the banner behaved normally.
Tuning was the whole job. A flat 400ms delay cut accidental taps sharply on small Android screens but also cost real clicks on iPhones, where the banner sat comfortably within reach and users were deliberately tapping it. So the delay was made conditional. The responsive site detects screen dimensions automatically, and the delay scaled with viewport width: 350ms below 380px, 250ms between 380px and 420px, and 120ms above that. iOS devices got the shorter end of the range because the layout there left more clearance around the banner edge.
Two details mattered more than the raw numbers. First, the overlay only blocked taps that arrived within 60px of the banner’s outer edge, since that is where stray thumbs landed. Dead-centre taps passed through immediately. Second, the delay applied only on the first render of the mobile lobby, not on every scroll back into view, so returning users were not punished.
What the delay did to stray taps and bonus clicks
Over the following three weeks, accidental taps on Android devices with screens under 400px wide fell by 61% against the pre-fix baseline. Welcome bonus clicks settled at roughly 11% above the original control, down from the 15% peak but still a clear gain. On Casushi iPhone the change was close to neutral: a 2% dip in bonus clicks, well inside normal variance.
The quality signal was the interesting part. Welcome bonus clicks that went on to complete a cashier deposit rose from 9% to 13% of total banner clicks. Fewer taps, better taps. Support tickets mentioning “the site opened the offer when I didn’t touch it” dropped to near zero in the same period.
A short delay is cheap to ship and easy to reverse. If you run a promo banner above a game grid, start with a 250ms edge-only block, watch your small-screen Android segment separately, and judge the result on deposit-qualified clicks rather than raw click volume.
What the Casushi casino app case means for banner and promo placement
The test result is less interesting than what it forces you to admit: a banner is not a static asset, it is an interaction surface that behaves differently on every handset. Apple users download the Casushi app from the App Store and get a fixed, predictable viewport. Android users access the Casushi site through a mobile browser, and that browser might be a 5.5-inch budget phone with a 720×1280 panel or a 6.7-inch flagship. The responsive site detects screen dimensions automatically, but a banner placed above the mobile lobby grid still sits within thumb reach of the scroll gesture. Treat placement as a device-aware decision, not a design preference.
Tap-target spacing matters more than colour, copy or animation. The accidental taps in the test clustered in the first 300 to 400 milliseconds after the lobby rendered, which tells you the problem was timing plus proximity, not persuasion. Keep at least 8 pixels of dead space between a promo banner and any adjacent game tile or navigation control, and never let a banner’s tappable area extend to the full width of the screen edge on small Android devices. The mobile lobby includes the welcome offer, so the banner competes with real content for the same thumb. Space it accordingly.
Push notifications deserve the same discipline. If the Casushi app fires a promo alert and the user lands directly on a banner that is already mid-animation, the first tap often misses. Deep-link to a stable landing state, let the page settle, then expose the call to action.
Here is the checklist that came out of the case, and it applies equally to the Casushi mobile site and the Casushi casino app:
- Segment banner placement by screen dimensions, not by device brand. Test at 360×640, 390×844 and 430×932 as a minimum before shipping.
- Hold the banner non-interactive for 300 to 500 milliseconds after render on screens under 400 pixels wide, and measure stray taps per session as a guardrail metric.
- Report welcome bonus clicks alongside a quality signal: deposit conversion rate, cashier deposit completion and bounce-back rate within 10 seconds.
- Keep one primary promo above the fold and move secondary offers below the game grid. Two competing banners halve the attention each receives.
- Re-run the test after any change to the cashier deposit flow, the welcome bonus terms or the app shell, because all three shift where users tap next.
The Casushi app supports Face ID logins, which shortens the path from open to lobby and compresses the window in which a user decides what to tap. That makes early-screen clarity more valuable than a louder banner. The Casushi app cashier manages payments, so a click that reaches the bonus page and then stalls at deposit is a failed click, however good the click-through rate looks in a dashboard. The Casushi mobile site offers responsible play tools, and promo placement should never crowd them out of the visible area; burying a deposit limit behind a banner is a bad trade in any jurisdiction.
One last point on distribution. Casushi does not publish a standalone Android APK file, so any banner promising a direct download is a trust problem waiting to happen. Point Android users to the browser experience and iOS users to the App Store listing, and keep the copy honest. This website operates as an information page only, so the placement lessons here are observational: measure stray taps, segment by screen, and judge a banner by the deposits it finishes, not the clicks it collects.
If you run any kind of promo placement on a mobile lobby, copy the sequence from this case rather than the outcome alone. Start by instrumenting clicks on the welcome offer banner before you move anything, so you have a clean baseline to compare against. Split the numbers by device class, at minimum separating Casushi iPhone traffic from Casushi Android traffic, and log screen dimensions alongside each tap. Then move the banner, wait for a full test window of at least two weeks, and compare bonus clicks and stray taps side by side. The 15% lift only meant something because the accidental-tap rate was measured in the same pass; without that second metric the change would have looked like a clean win when it was not.
For your own build, the practical next step is to add an interaction delay to any banner that sits close to the game grid and tune it per screen width. A short pause of roughly 300 to 500 milliseconds on smaller Android viewports, paired with a tap target of at least 48 pixels and clear spacing from the first row of tiles, is a sensible starting point. Test it on a real handset rather than a desktop emulator, since the responsive site detects screen dimensions automatically and the layout you see in a browser window will not match a 360-pixel-wide phone. Keep the cashier and responsible play tools a single, obvious tap away, and avoid push notifications nagging users toward the bonus. Finally, note that this website operates as an information page only and does not take bets; for the app itself, Apple users download the Casushi app from the App Store, while Android users access the Casushi site through a mobile browser, because Casushi does not publish a standalone Android APK file.