App pages on the App Store and Google Play Store share the same combination of elements – an icon, description, screenshots gallery, reviews, and sometimes a video. However, the two major app stores offer different layouts, which causes differences in the significance of those elements to users.

ASO Experts at Lab Cave were aware that the platforms work differently and vary in the type of audience.  They had been A/B testing their Google Play visuals for a while until the time came for App Store screenshots.


It has never been easy to allocate a large budget for acquisition campaigns just to test icons or screenshots. However, as the company grew, we took a step forward and decided to optimize conversions for our games on the App Store.

Yuki López, CRO Specialist at Lab Cave

To bring this into action, Lab Cave turned to the Splitmetrics A/B testing tool.


App Store screenshots: when order matters

Lab Cave tested the Viking: Heroes War” app page screenshots with SplitMetrics.   

Discovering what works better for different app stores was a strong reason, but the main one, however, was to identify what boosts the conversion rate from page visits to installs. Although at that time the icon on the game page had a high Tap-Through Rate (TTR), the conversion from page visits to installs was low, compared to other games of the same kind.

On the Play Store product page, the image gallery starts with the game cover, which has performed quite well for Android users. When designing App Store screenshots, Lab Cave agreed that putting the game cover as the first image would also be a good idea. This, however, required verification with A/B testing.  

The Lab Cave team carried out market research for the game to find out that users may also want to have a glance at how the game is played. As a result, the game cover was combined with the actual gameplay screenshots, and different screens order was tested, with all other things being identical.

viking heroes control
viking heroes screenshot variation
viking heroes app store screenshots
viking heroes optimized screenshots

The results revealed that the control version (banner-like game cover + gameplay screenshot) performed better than the others. In addition to having the highest Click-to-Install Rate (23.5%), the control version showed the lowest bounce rate of all the sets.

Additionally, it was insightful to find out that even though the last set of screenshots didn’t win in the test, it demonstrated the biggest number of direct installs.

App Store screenshots: brighter, simpler, banner-like

Another attempt to discover a booster for a visit-to-install conversion rate was A/B testing for “Afterpulse – Elite Army”.

The hypothesis was that render-based screens and CTAs with direct messages attract users’ attention better. We also wanted to test if brighter colors lead to a bigger number of installs.

Yuki López, CRO Specialist

The team set their minds to optimizing the App Store screenshots for the game. They tested render-based against gameplay styles, introduced brighter colors to some of the images and simplified CTAs.

Afterpulse app store screenshots
Afterpulse winning image gallery

The findings proved the hypothesis. The winning variation included an outstanding first banner-like screenshot with the game title included. The set brought a 17.8% increase in conversion against the control version. In fact, the winner showed the highest Engagement Rate (86.9%) and the longest Average Time on Page (14.11 seconds). The results are supported by an impressive confidence rate of 98.5%.

App Store screenshots: when banner-like is better (again)

For “Tap Knights: Heroes & Monsters”, Lab Cave carried out market research to find out that the horizontal orientation for App Store screenshots would probably do better. All SplitMetrics experiments included landscape images, but what they wanted to test in the first place was the style render-based against gameplay-based.

The hypothesis was simply that render-based horizontal screens would perform better for this genre of games.

Carlos Bancalero, CRO Manager at Lab Cave
tap knights app store screenshots
tap knights winning images

Optimization yielded a 45.8% improvement compared to the control version. The results supported the hypothesis, showing us (with a 98% of confidence) that render-based horizontal screenshots perform better than gameplay based ones.


By getting their game apps A/B tested, Lab Cave expected to identify what kind of visuals work better for iOS users as well as to get a deeper insight into user behavior on product pages.

In all 3 tests, screenshot 1 of the winning variations was banner-like. In general, render-based images are powerful attention grabbers. They perform even better when placed at the beginning of the image gallery, as they demonstrate the idealized look and feel of the app.

With the help of SplitMetrics, Lab Cave managed to compare user behavior data for their game (Viking: Heroes War”) in two major stores. In this particular case, the tastes of Android vs iOS users didn’t differ: the set starting with the game cover performed the best for the App Store and the Play Store.

Overall, numbers speak to the success of App Store screenshots optimization – 17.8% and 45.8% increase in conversions from page views to installs.

In general, A/B testing has shown us that sometimes personal opinions about visuals can be completely wrong, and that gathering data through tests is essential to succeed in this business.

Carlos Bancalero, CRO Manager at Lab Cave


Want to achieve similar results?
Schedule a demo call with one of our optimization experts
Schedule Demo