Overview of Smart Segmentation
Smart segmentation enables you to identify and monetise app users who are unlikely to spend in your app. Smart segmentation uses machine learning to predict your users’ purchasing behaviour and segments them into two groups:
- Predicted non-purchasers: Users who aren't likely to make in-app purchases.
- Predicted purchasers: Users who are likely to make in-app purchases, like buying extra lives, etc. Limiting the amount of ads shown to these users may help to improve their overall in-app experience and lifetime value.
With smart segmentation, ad requests for predicted non-purchasers will be filled, while ad requests for predicted purchasers won't be filled. This may help to unlock new revenue by monetising the non-purchasers with ads, while preserving an ad-free experience for the purchasers.
On average, predicted non-purchasers make up about half of your users.
Note: Behind the scenes, smart segmentation predicts the likelihood that a given user is a non-purchaser and alters ad serving accordingly. When a user is ineligible for personalised ads or has disabled them, Google doesn't predict purchaser likelihood. Accordingly, users who don't receive personalised ads will experience the same serving behaviour as predicted purchasers.
Smart segmentation is only available for interstitial and rewarded ad units.
For interstitial ad units, there are two ways to use smart segmentation:
- Experiment mode: Experiment mode allows you to test the ad unit with smart segmentation enabled on 10% of your app’s users for 90 days. Learn more.
- Full mode: The ad unit with smart segmentation enabled is used on 100% of your app’s users. Learn more.
In your Home dashboard, the smart segmentation card lists your apps that aren't already using the ad format and are eligible to use smart segmentation. App eligibility is based on user traffic and the amount of in-app purchases visible to AdMob. Additionally, iOS apps must be linked to Firebase to be eligible for smart segmentation.
The smart segmentation card estimates the increase in ad earnings that a smart segmentation interstitial or rewarded ad unit could provide when used in addition to existing ad units. Note that replacing existing ad units with smart segmentation ad units can result in lower revenue and impressions as smart segmentation ad units only show to a portion of your users.
You can monitor the performance of your smart segmentation ad units in your App overview dashboard. To view your app overview dashboard, click Apps in the sidebar. Select the name of your app, then click App overview in the sidebar.
You’ll be able to measure the progress of the smart segmentation ad units with the following metrics:
- Est. ad earnings: The estimated amount of revenue generated by the smart segmentation ad units.
- eCPM: An estimate of the revenue that you receive for every thousand ad impressions for smart segmentation ad units.
- Impressions: The total number of ads shown to users for smart segmentation ad units.
- Model accuracy: The percent of purchasers who were accurately predicted by smart segmentation.
- Non-purchasers: The percentage of users that smart segmentation predicted to not make any in-app purchases. These users are shown ads.