A/B test is one of the basic ways to check ideas in mobile applications on a rapid scale. The arrangement is usually clear: you can take a test for a few weeks, measure its effect, and decide to offer more feature or repetition.
One of the main aspects but that is usually ignored to test A/B is to determine the correct exposure points. In this article, we will cover the reason for preparing the correct exposure point, common risks to avoid it, and providing some tips and tricks that you learned along the way.
What are the points of exposure, in any case?
The exposure point is the real moment that the user faces at first or interacts with the feature it tests. For example, the timing when the user sees a new button, or when they see an intended page that is re -designed after clicking on something.
Why is it important to choose a good exposure point?
If you are staring at the results of A/B testing and not logical, bad exposure points may be the perpetrator. Bad exposure points can lead to:
- Conjried data: It will not be clear whether your features are a blow or miss.
- Hidden insects: Everything may seem well on the surface, but serious problems such as the application of the application may slip under the radar.
- Lost opportunities: Your data may not appear wrong with any major effect, although users have already enjoyed the feature.
An example of a bad exposure point
Imagine you are testing Upsell for the new subscription in your application, but you are raising the point of exposure to running the application instead of users already displaying the subscription page.
problem: Perhaps only 10 % of users see the UPSell page, making 90 % of the test data without value for this decision.
Loose against narrow exposure points
Loose exposure points:
This happens a little very early. Users are experienced before testing the feature that has been tested. This early exposure spoils the data, and it is difficult to find a real effect.
Narrow exposure points:
These happen at the exact moment that the changing users tested. Data collected with more accurate and reliable exhibition points, and easier to analyze.
Which is better?
It depends on your use. Narrow exposure points are preferred because they provide you with cleaner and more specific data, even with smaller sizes. Sometimes, however, narrow exposure is not possible. In this case, you can use loose exposure points with knowledge that is likely to require a larger sample size to achieve important results.
Avoid accumulating changes
Do not mix or pile of multiple exposure points together in one test. Divide each change into its A/B test. Although it will take a little longer, your data accuracy will be much larger, and you will have more accurate conclusions about each individual feature.
A quick example of life
Suppose you add a new row type in table width:
- Good exposure point: Detailed exposure when the new row is presented in opinion.
- Bad exposure point: Exposure to fire once the table is loaded, even if the new row is not visible yet.
Final advice
- Plan for the future: Think with your exposure points at the time planning time.
- Repeat quickly: If the results of the test are not as expected, it will be repeated quickly.
I would like to hear your stories and experiences in determining your A/B testing points. Share it in the comments!
