Mobile App A/B Testing tips

When you start working with mobile apps, you can often hear an unknown phrase “mobile application A/B testing”. Am I right? Well, there’re no surprises, then.

The thing is, there are 1.6 million apps for Android in Google Play and 1.5 million iOS apps in Apple App Store as of July 2015. So, obviously, if you create a new one, you need to be prepared your company will have a lot of business competitors.

There are lots of questions you should answer before your app’s release. For example, which icon is better, what should be written in the description, in what order screenshots should be placed. Even what should be the color of the icon’s background. You can’t even imagine how many little things have an influence on the final decision.
Mobile app a/b testing tips
Many marketers do researches, analyze competition and best practices and go with their guts, surely. But who can guarantee that the very same methods worked for competitors will work for you? Maybe your audience has different opinion on usability.

Well, then you need to test your exact product. And that’s where comes mobile application A/B testing.

But what is A/B testing at all? The definition is the following: statistical hypothesis testing with two variants. Sounds interesting? Let’s begin then with our mobile app a/b testing tips!

how to run mobile app a/b testing

Main problem with App Store – it doesn’t allow you to perform A/B testing for mobile apps, unlike Google Play. So for iOS A/B testing we need to mimic App Store experience.

You should create an A/B landing page, send there traffic and analyze, simple as that. You can use different template tools for the landing page, create it by yourself or ask any developer, for example, us.

Narrow your choices

After you create a landing page, you need to choose two main variations of the product you test – icons, for example. They should be pretty similar, but slightly different. For example, you create two icons with a different background, and test them.

Differences may be very small, but even the simplest detail can have an influence on user’s decision.

Imitate for your users two landing pages, similar, just with different icons, as in our example. It can be anything, from the description to price for upgrade to premium version.

Main rule of native mobile apps a/b testing is – you need to change only one thing, you test for. There can be one change, that’s it.
Native mobile apps a/b testing

find the traffic

For testing, you need to find some users who are interested in your app. Or make them interested, doesn’t matter. You can start, for example, some pre-ordering campaign for signing up, for example. Obviously, we can’t count on the amount of the downloads as we haven’t uploaded the app to the App Store yet.

So you need to find your own conversion rate, what it will be for you. It can be amount of clients signed up for emails for the future release or amounts of clicks, it’s up to you.

You can use AdMob, Google Adwords or any other mobile ad network to run this small campaign. You can use a network like AdMob for a quick and easy set up, but a more targeted network like Google Adwords can have its advantages, too. In that way you will ensure your test will be found by exactly your auditory.

Read also:

find out the results

Finally, to the main thing. Wait for at least 100 clicks so your statistics will be clearer. If you create landing page with help of custom code or templete, you also need to use some service like KISSmetrics or Optimizely to track the number of clicks.

Obviously, your final result is not showing you exact success of your app, how sales will go and other things. But you’ll get a general idea, understand what choice was liked by bigger amount of the users – that is your desired option for the future use.

The way to increase mobile app sales with a/b testing is to constantly run it, make improvements to existing product. If you make UX even a little bit easier, it can do magic. Like, really.

And, of course, don’t forget, no matter what, your product can be different for iPad and iPhone users, for example, so results and marketing depend on that also.
Find out the results of a/b mobile testing


So here we showed some of mobile a/b testing tips. There are lots of them, and all are individual, depending on the every project, of it’s main purpose and other things.

A/B testing of course isn’t the rescue, but you need to give it a try. At the very least, who doesn’t want couple of thousands downloads more?

If you have any questions regarding A/B testing, creating a landing page or general developing – write in the comments bellow or contact us here, we will gladly help you!