We’ve Acquired data.aiLearn more

Blog

APP STORE OPTIMIZATION · HUGH KIMURA · JULY 2013

How To A/B Test Your App Screenshots To Increase Downloads

Learn how to improve the number of downloads your app gets by doing A/B testing.

As we mentioned back in May, your app screenshots can be a big influence on if your app gets downloaded or not. This is especially true of your first screenshot because it is the first thing that people see when they find your app through search. Here is an example of a great first screenshot.

lt="Real Racing screen example

In this post, you will learn how to setup a simple A/B test to determine which of your screenshots is more appealing so you can use that screenshot first. Since we cannot A/B test in the App Store, we will have to recreate a similar experience on a website.

Set Up A Mobile Landing Page

A landing page is just a single page that you will send visitors to, in order to see which screenshot they like best. You can create the page yourself or use a responsive theme. Sites like ThemeForest will sell you a theme that is ready to go for a few dollars.

There are also mobile landing page builders like Landr, Convrt or Pijnz that you can use to create a mobile enabled page in a few minutes. It might also be worthwhile to hire a designer on oDesk or Elance. Then download the App Store badges for people to click on.

lt="Landing pages that help you experiment

Create An A/B Test

Now that your website set up, create two different landing pages with the two versions of your primary screenshot. You can use Google Content Experiments, Optimizely or Visual Website Optimizer to conduct your A/B test. Simply insert the testing code into your pages and your variations will automatically be served randomly.

lt="Testing app visual elements

Get Some Traffic To Your Page

There are several ways that you can get visitors to your landing page. You can send traffic from social media, an email list, or you can purchase traffic from sites like Google Adsense or Flurry AppCircle.

lt="Landing page tools

Purchasing ads is probably the best way to test your screenshots because most visitors will not already be familiar with you, like with social media or your email list. As people come to your site, they will "vote" for a screenshot when they click on the download button. If you pay for traffic, your test could cost as little as $75.

Analyze The Results

Then take a look at your results and see which version of your screenshot got more clicks. Be sure to get between 100 and 200 total clicks before you finally declare a winner.

If you would like to create a multivariate test with more than two screenshots, you can do that too. Remember that the more variations you have, the more traffic you will need to determine a winner. Try to keep the number of variations to as few as possible, and only test the best candidates.

Conclusion

By using this method, you can cheaply test the appeal of your first app screenshot. It is not a failsafe method, but it is much more conclusive than guessing, or relying on the opinions of your friends. After you do this test, do not forget to upload different screenshots to the other screenshot 4 slots, for maximum visibility.

Have you A/B tested your screenshots before? If so, what were the results?


Sensor Tower's platform is an enterprise-level offering. Interested in learning more?


Hugh Kimura

Written by: Hugh Kimura, Head of Content

Date: July 2013