How to optimize and quickly increase the number of app store users?

On my first day at CallApp, the CEO gave me an extremely attractive challenge. He came over to me and said, “We need to do everything we can to increase the number of downloads per week (an acceptable measurement unit).”

After receiving the task, the first thing I did was to analyze the current download volume, hoping to find out which aspects to focus on to achieve the maximum improvement effect. After a lot of data processing, I decided to focus on improving the conversion rate of Google Play Store.

I will explain to you why I chose this path? Can you also explain some of the knowledge related to app store conversions ?

There are many options for downloading apps, but for Android users, the most common way is through the Google Play app store.

In the Play Store, users may search for app names or keywords . The search results may have a large number of matching apps (if searching by keywords ) or the apps themselves are listed. The user then needs to choose which one they want to download and then download it. Our goal is to convert searching for apps into downloading apps. So we can infer that improving the search process can increase our downloads.

Google provides various testing metrics for this purpose in the developer dashboard. With so many different metrics out there, I had to choose which one to use to help me increase downloads.

The indicators I initially selected were:

• Path to store — user acquisition (paid or free traffic );
• Store conversion rate;
• Login process – Registration wizard;
• App Retention – Increase app usage (acquired through user end).

One thing to note is that the above indicators do not require the same quantitative source. And one more thing to consider is that they have completely independent time frames (cycles).

For example, "login" or "new features" require a lot of development costs (releasing a new app version). If there are other tests that need to be done at the same time, this "big move" will result in more time. In fact, I only need to choose between long-term testing and fast store ranking testing, and the latter only requires a good designer and a product manager with an open mind.

To be more specific, the status of our company's app at that time was:

• Conversion rate (installs) on the Play Store page -25%;
• Registration conversion rate (login) - 50%;
• 7-day retention rate -30%.

To summarize:

• Improving store conversion rate can significantly increase the number of users (with the best improvement effect); the translator and reviewer unanimously agreed that the data in the author's figure is obviously problematic, and the improvement effects of the four methods are actually the same, all 13%!
• No development costs required: It is easy to test and operate through the indicator tools provided by Google;
• Making changes without releasing a new version is the fastest way to test and change your overall conversion rate.

In the App Store test, the following indicators can be tested:

1. Visual features: The main visual content displayed on the page, including: pictures, text/information, video - yes or no.
2. App icon: icon design.
3. App name: Title (first 30 letters).
4. Description: A brief description of the app (first 80 characters).
5. Full description: A long introduction about the app.
6. Screenshots: Includes: app loading screen-pictures, and the order in which the screenshots are presented.

Therefore, all these tests can be based on the current online version, and the tests and results can be operated and obtained through the Google Play developer panel.

This makes it possible to test each element individually, and you can also test the impact of different elements on users of different languages. There are other factors that influence whether users choose to download the app, but testing these factors requires releasing new versions to achieve and produce results.

Practical example: CallApp ranking test in the store

In the second quarter of 2016, we ran about 60 tests on the App Store, processing and analyzing every element. Finally, 20% of the experimental results were applied to our app’s listing display in the Google Play store.

Here’s how to do it:

1. Only one parameter was varied in each test;
2. Conducting tests in different languages ​​(to analyze the impact of language on different cultures);
3. Even if the test is successful the first time, we will repeat the test to ensure the validity of the test results.

The subjects of the test were classified as follows (60 times in total):

Types of practical tests (15 in total):

Test case: App Icon

We tested more than 20 different app icons. There are three places where users might see an app icon:

1. App list in Google Play Store;
2. Application home page;
3. After downloading, on your mobile screen.

App icon testing is a sensitive test that can discover many far-reaching results. Any changes in the icon must reflect the company's marketing strategy, and it will also affect the company's overall brand image. Therefore we decided that any changes we made to the icon would have to be as minor as possible to minimize impact on the company brand.

Under this premise, we made the following plans for App icon testing:

• Test different icon shapes;
• Test different icon colors and backgrounds;
• Test 2D and 3D effects;
• Test simple icons;
• Add elements to the icon.

In order to test an optimal value, the improvement should be significant enough (you can clearly know on the control panel which variable has the best effect).

This is the icon we originally put in the store:

Test 1 :

Three variables. We assume that users may be more familiar with circular icons, so we added three different colors of circular backgrounds to the original icon. At the same time, we tried to make the icon glow against the background. This didn’t work very well!

Test 2:

Next we try to add a 3D icon on the circular background. The result: the metrics showed improvement, and the new icon was implemented in the production version.

Test 3:

We wanted to see if a square shape could work better than a circle, so we gave the icon a square background. This has no effect!

Test 4:

Then we went back to the circle and tried to make some changes to the circle icon, but this didn’t work!

Test 5:

We thought maybe changing the color of the icon itself would be a good idea, so we tried designing a colorful icon that would stand out among other apps.

no! It didn’t attract our users.

Test 6:

We then tried to test adding a blue background to the 3D circular icon, because many apps now like to use a blue background. We also made some shape separations.

no! The idea was unsuccessful.

Test 7:

We try to elevate icons with little touches like “first place” badges. After all, everyone loves to be number one, right? Totally useless!

Test 8:

Our final attempt was: we added a white border to the icon style of Test 2.

The winner has emerged! This design brought better feedback than expected.

Conclusion

Although there seems to be no clear conclusion to tell you which one is useful? Which one is not okay? But through experiments you can use data to know which one works and which one doesn't.

What I want to tell you is: while you are doing some innovation, you should also continue to test your ideas, no matter how big or small the idea is.

Is there anything that hasn’t been mentioned before? some.

Try to run separate experiments by country rather than language, because version optimization and localization are based on countries. For example: Many countries speak Spanish or French, but the cultures between these countries are very different.

Finally, two useful suggestions for implementing experiments:

1. The first is to measure only one variable per test;
2. The second is to conduct tests based on assumptions/hypotheses.

While Google provides you with great metrics tools, don’t forget to familiarize yourself with the methodology. Even something as simple as A/B testing can go wrong, especially if you don’t follow the rules.

The author of this article @王猫猫’s name is taken. It is compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting!

Product promotion services: APP promotion services, advertising platform, Longyou Games