Tutorial: Optimize AdMob ad frequency

Step 2: Set up an A/B test in the Firebase console


Introduction: Optimize AdMob ad frequency using Firebase
Step 1: Use AdMob to create new ad unit variants for testing

Step 2: Set up an A/B test in the Firebase console

Step 3: Handle Remote Config parameter values in your app's code
Step 4: Start the A/B test and review the test results in the Firebase console
Step 5: Decide whether to roll out the new ad format


Now that you have new ad units implemented in your app, you need to set up an A/B test that will help you understand how these ad units perform. You'll define which users to target in the test as well as your testing goals. You'll also need to define your test variants (the different ad units in your app) and set up the Remote Config parameter which will control the display of the variants in your app.

Firebase A/B Testing utilizes the following products to test and analyze the effects of adding an ad unit to your app:

  • Firebase A/B Testing (this step) — define goals and configurable parameters for your test
  • Firebase Remote Config (next step) — add logic to your code to handle the configuration of the parameters
  • Google Analytics (runs behind the scenes) — measures the impact of the configurations

Initiate a new A/B test

To initiate a controlled test for optimizing ad frequency, start by navigating to the A/B Testing section of the Firebase console. Click Create experiment, then select Remote Config.

<span class=Firebase console UI showing how to start an A/B test using Remote Config" class="screenshot">

Set up the basics

In the Basics section, define the experiment name and provide the experiment description.

<span class=Firebase console UI showing how to set up the basics of A/B test" class="screenshot">

Set up targeting

  1. In the Targeting section, select the iOS or Android app that the experiment will target.

  2. Set the percentage of users who will be exposed to the experiment. For this tutorial, the new ad units will be tested with 30% of your users. Note that this doesn't mean that 30% of all your users will see the new ad units with increased frequency caps; this means that 30% of your users will be exposed to the two new interstitial ads along with the baseline ad unit (your existing ad).

    Leave all other settings as their defaults.

<span class=Firebase console UI showing how to set up the targeting of A/B test" class="screenshot">

Set up your goals

Firebase A/B Testing tracks a primary metric to determine the winning variant, but it also allows you to add secondary metrics to understand the impacts of different configurations on other important factors for your app.

  1. For this tutorial, Estimated AdMob revenue optimization is the primary goal, so select it from the dropdown menu.

  2. (Optional) If you want A/B Testing to track additional metrics, like Estimated total revenue or different retention rates, select those by clicking Add metric.

<span class=Firebase console UI showing how to set up the goals of A/B test" class="screenshot">

Set up the variants

The last step of configuring an A/B test is defining a Remote Config parameter that controls which ad unit will be shown to users.

  1. In the Variants section, create a new parameter named INTERSTITIAL_AD_KEY by typing it in the Parameter field of the Baseline card.

  2. Finish setting up the Variants section using the following settings for the INTERSTITIAL_AD_KEY parameter:

    • Baseline variant: Value set to the ad unit ID of the existing ad unit with the frequency cap setting of 4 impressions per user per 10 minutes
    • Variant A variant: Value set to the ad unit ID of the new ad unit with the frequency cap setting of 6 impressions per user per 10 minutes
    • Variant B variant: Value set to the ad unit ID of the new ad unit with the frequency cap setting of 8 impressions per user per 10 minutes

    For this tutorial, the Baseline variant will show the existing ad unit (4 impressions per user per 10 minutes). Variant A and Variant B will show the new ad units (6 and 8 impressions per user per 10 minutes, respectively) to a small subset of users. This is controlled by the parameter's value which is the ad unit ID taken from the AdMob UI. These parameter values are set here in Firebase A/B Testing, but it's actually Firebase Remote Config that sends these values to your app's code for handling. You'll set up Remote Config in the next step.

<span class=Firebase console UI showing how to set up the variants of A/B test" class="screenshot">

Click Review to make sure your experiment is set up as expected. However, before you can actually start the experiment, you need to define how your app's code will react to the parameter values received from Firebase. Proceed to the next step to implement how Remote Config handles the INTERSTITIAL_AD_KEY parameter.



Step 1: Use AdMob to create new ad unit variants Step 3: Handle Remote Config parameter values