Conversion rate optimization is one of the most overlooked yet critical marketing tools. It’s the most effective way to get more users, lower your user acquisition costs and amplify all of your marketing activities without blowing up your budget.
A/B testing (aka split testing) is a process of comparing two or more versions of your app page to find out which one performs better.
The process is pretty straight-forward:
Step 1: Concept
You come up with a concept of how you can tweak your App Store or Google Play page to increase its conversion rate. Here’s how you can generate ideas for experiments:
- Look at your competitors and apps from other categories for inspiration
- Check out case studies of successful A/B tests
- Follow industry best-practices
Step 2: Setting up Experiment
Apple App Store: there’s no way to run A/B tests directly on the App Store, therefore you need to create different web-pages with your variations that look like App Store pages and send users there. Best way to do this is via a specialized A/B testing tool: Splitmetrics.com is a great option with a free trial.
Google Play: for Android apps you can either opt for Google Store Experiments - a solution from Google that lets you run A/B test ideas right in Google Play. However if your app is not yet live, or you need to control which user segments are in the test, and get advanced analytics you will need a specialized tool like Splitmetrics.
Step 3: Running the Experiment
Once your variations are set up you need to drive traffic to them. You need a source of quality non-incentivized traffic to fuel your A/B tests: Facebook ads is one of the best options for that.
Make sure that you only send qualified users (e.g iPhone users for an iPhone app A/B test), matching the language of the landing page, etc. Narrow down your audience by targeting people who match your target user profile. Run your A/B test over an extended period of time (at least a full weak) to help exclude fluctuations.
Step 4: Analyzing results
As soon as your test reaches statistical significance with a high confidence level you can finish it and start analyzing the results. When there is a winner (i.e. conversion rate of one variation is higher than that of the others) you can push it live to the App Store or Google Play.
If there is no clear winner - analyze the secondary metrics: time on page, bounce rates, engagement, etc. See how you can build on the improvements and do better in your next A/B test.
Additional Points to Keep in Mind for Effective A/B Testing
1. Test one element at a time
This way you will know exactly what changes impacted your conversion rate. If you introduce too many alterations at a time, only some of them might have positive effect and you’ll end up guessing which ones.
App Page Elements to A/B Test
- Video Preview (and poster frame)
- Short Description (before the ‘...read more’ link)
2. Don’t project App Store results to Google Play and vice versa
There are just too many differences between users, devices and store layouts, so you can’t just take your Google Play A/B test winner and push it to the App Store. You need to test separately for the two platforms. Otherwise you may end up hurting your conversion rates.
3. Traffic is key
Knowing exactly who your desired and most valued users are and targeting them in your A/B tests is a critical component. This way you will not only increase your conversion rate, but you will increase it where it matters - so that it helps acquire the right type of users.
Even a slight increase in conversion rate will yield enormous result in the long run saving you budget by reducing UA costs as well as helping better understand your users. If you manage to make A/B testing an integral and recurring marketing activity it will ensure sustainable growth and put your app ahead of the competition.