What is ad frequency optimization?
Whether your app is hybrid-revenue or ads-revenue driven, optimizing ads revenue and keeping a high-quality user experience can be tricky. Ads are a great source of revenue, but a high frequency of ads can provide a negative user experience and might lead to user churn.
There is no "one ad frequency suits all" approach for any app; ads performance varies greatly from app to app and from audience to audience. You might be concerned that increasing ad frequency could have a negative impact on user experience or retention, but you might also be curious to see if it could lead to an increase in revenue and engagement when instrumented properly, keeping engagement metrics in check.
To resolve these unknowns, Firebase offers tools that help you test and then make data-driven decisions about the optimal ad frequency:
Using Firebase, you can A/B test the performance of various ad frequencies with a small subset of users.
You can observe the test results and review recommendations from Firebase about which ad frequency is performing better and with minimal impact on retention.
Once you're confident that the changes will likely have a positive impact, you can roll out the changes to more of your users with a click of a button.
Business case and the value
Developers and publishers using Google AdMob and Firebase tools for optimizing their ad frequencies enjoy major revenue uplifts without adversely impacting user experience.
Qtonz uses Firebase to boost ad revenue by 4x and grow engagement by customizing the experience for different stages of the user journey.
Implementing the solution
To implement this solution, you can follow our step-by-step tutorial (find an overview of this tutorial later on this page).
In this multistep tutorial, you'll learn how to use Firebase to test various frequency caps for Google AdMob ads in your app. It uses interstitial ads as the example test case, but you can extrapolate and use these same steps to test frequency capping for other ad formats.
This tutorial assumes that you already use AdMob in your app and that you'd like to test whether changing the frequency of an interstitial ad unit will have an impact on your app's revenue or other metrics. However, if you don't already use AdMob in your app, that's ok! The steps in this tutorial can also help you understand what ad frequency you should use in your app.
Products and features used for this solution
Google AdMob enables you to create ad units with various ad frequencies or refresh rates that will be served within your app. When you link AdMob with Firebase, AdMob sends ad revenue information to Firebase to improve ad strategy optimization.
Google Analytics gives you insight into user engagement, retention, and monetization metrics like total revenue, AdMob revenue, purchase revenue, and much more. It also allows you to create user audiences and segments.
Firebase Remote Config
Firebase Remote Config enables you to dynamically change and customize the behavior and appearance of your app for desired user segments — all without publishing a new version of your app. In this tutorial, you'll use Remote Config parameters to control which ad unit is shown to your users.
Firebase A/B Testing
Firebase A/B Testing provides the interface and infrastructure to run product and marketing experiments in your app. It takes care of distributing experiment variants to users, and then performs statistical analysis to determine if an experiment variant is outperforming the control group based on your selected key metric, such as revenue or user retention.
Solution tutorial overview
Go directly to the step-by-step tutorial
Use AdMob to create new ad unit variants for testing
Create two new interstitial ad units in AdMob.
Set the Frequency capping of each ad unit to an impressions per user value that you want to test.
Implement the ad unit placements within your app's code.
Set up an A/B test in the Firebase console
Define testing basics, targeting, and the goals that the test will run against.
Define test variants and set up the Remote Config parameter that will control which ad unit is shown to users in the test.
Handle Remote Config parameter values in your app's code
Use the Remote Config parameter in your app.
Implement the logic for displaying the ad unit based on the parameter's value.
Start the A/B test and review the test results in the Firebase console
After starting the test and allowing it to run for a few days or weeks, check the Firebase console for whether the A/B test has a winning variant based on the primary goal of the A/B test.
Review the impact on secondary metrics for each variant to ensure the variants didn't cause unintended negative impacts to those metrics.
Decide whether to roll out the new ad unit with the updated ad frequency
If A/B Testing determines that the variant showing the new ad format is the winner, you can start showing the ad format to all users targeted in the experiment, all users of your app, or to a subset of your users.
If a clear winner isn't yet determined, you can either continue running the experiment to gather more data, or end the experiment if it's already been running for a long period with inconclusive results.
View a list of common terms for this solution
View a list of common terms for this solution
AdMob revenue: AdMob network and open bidding revenue
IAP revenue: In app purchases revenue
Total revenue: Total revenue
Retention: Retention as a key metric in A/B tests is tracked as 1 day, 2-3 days, 4-7 days, 8-14 days, or 15+ days user retention
Remote Config parameter: The configurable parameter used to control which ad unit is show to users. In this guide, it will be an ad unit ID.
Baseline configuration: The as-is configuration in any particular A/B test — also known as the control. The control usually uses the default value for the Remote Config parameter, but it can be configured to use a new control value if needed.
Variant configurations: The variant configurations are the alternative configurations with different Remote Config parameter values that we would like to test against the baseline configuration.