As we previously discussed, In-App Purchase is becoming the primary source of app income. Aside from following implementation best practices, it is important to assess the performance of different in-app items. Assessment can come in the form of monetization KPI tracking or better yet, an A/B test.
The tracking of user behavior isn’t of much use unless you know how to act on each behavior. A/B testing allows us, marketers, to explore and experiment different monetization strategies with minimal risk to revenue loss.
IAP A/B Testing Best Practices
When is the right time to start an IAP A/B test?
It is advisable to start monetizing with the use of in-app purchases after the app has significant traffic and loyal user base. You can start an A/B test anytime you feel the need especially when you are still tweaking with the IAP monetization strategies. There are times where an A/B test is vital, though. This includes times when you conversion rates seem to be slow or skewed, if revenue is low, and if you add or change any in-app item.
Which IAP elements should you test?
It is important to consider first and foremost which stage of the monetization funnel directly affects the result you want to test for. Is the low revenue because of the IAP’s placement, price, or its very nature? Test or add IAPs at each level or part of the app where you feel the users are inclined to spend money on. The test would show if user preference over one item before or after an action.
You don’t need to specifically test every in-app item. There are in-app elements/aspects that you can tweak and A/B test for performance:
- Price
- Placement (or frequency of appearance)
- Design (color, layout, etc.)
- Headlines Wording
- Nature of the offer (bundles, boosters, extensions, etc.)
What types of test should you try?
You can try price point tests, placement tests, conversion funnel performance, and market segment test.
There are times when tests generate negative or no conclusive reports. This just means that you need to create and try new hypothesis or combinations.
How can you track the results?
App analytics tools are the first option. Google Analytics has a built-in A/B testing tool called Google Analytics Experiments. There are also other A/B testing tools that help in simplifying the testing process and offer special features that can help given an insight into what users want and need.
How long should the test last?
It is advisable to continue running the test until a statistically significant result is reached. That is if the aggregated data can be used for objective comparison. The longer the test runs, the more accurate the results will be.
But how would you know if an in-app purchase really won over the other? You can use A/B test duration calculators to calculate the amount of time necessary for the test. You can also try the traditional way wherein you just estimate the sample size you would need and then divide it by the daily traffic. The result would determine what sample size you should stop the test.
