Sometimes there are big disputes in developer teams on choosing a better app design for a store. Even specialists with huge experience make mistakes and make wrong choices. Is there any solution to the problem? Yes, A/B-testing. Today we’ll tell you how to carry it out with minimal expenses.

Basic principles

The essence of the method is to show variants to various user groups and analyse the acquired data. For example, you have several variants of app page design and you don’t know which one will work more efficient. So you reveal both of them and they are randomly shown to users. Then you have to calculate the conversion of each variant and compare the results. Such a testing solves important marketing tasks and often allows to significantly increase conversion.
The main conditions for getting statistically correct result:

  • ensuring representativeness of the selection.
  • using traffic from the same platform for each variant.
  • parallel measuring at the same time.

 

Testing process

It’s very difficult to carry out a full-fledged A/B-test in a store due to many evident reasons. That’s why landing pages are usually used in the web. The approach looks like this:

  • choose two or several variants of tested elements, for example, icons.
  • create totally similar landings for them.
  • direct traffic from the same source to them.
  • make an analysis in order to define the level of efficiency for each variant.

1

The secret of this approach is very simple: created landing pages are very similar to the store interface. They practically copy the app page on the market. Thus, users may not suspect that they aren’t in App Store or Google Play but in the web.
The method doesn’t require special knowledge and expensive tools, i.e. you can easily carry out A/B-testing. Tests are carried out in the web so you don’t have to work with a code, build in SDK or make an app update. You can test an app even before release with such an approach without wasting your time and potential users after app launch.
As for traffic, it’s better to attract it for testing from the sources you would engage during further promotion. In order to get authentic results, several hundreds of users is usually enough.

You can test:

  • icons.
  • name.
  • description.
  • screenshots and their order.
  • app price.

 

Tools

You can create landing pages that copy an app page in a store in different ways, even manually. But there’s an easier way — special services for testing app design. You can create landing pages that totally copy app pages in App Store or Google Play automatically there.

2
Test Nest and Split Metrics will help you to do this. The work scheme is quite simple: you only need to upload graphic materials (icons and screenshots) and enter text (name and app description). Later on, ready landing pages are placed in the Internet and then you can start an active stage of A/B-testing. Analytics data will be available in a convenient and comprehensible format.

This method of A/B-testing is rather flexible: you can check separate elements and app pages on the whole. Along with conversion coefficient, the most important metric, other ones are available, for example, time of page visits. There’s also a possibility to track types of user devices.

 

Conclusion

A/B-testing is the most important marketing approach which allows to choose the most efficient design elements with simple tools and without big expenses. The described approach will help to carry out testing before app release without loosing time and potential users.