Blog
In February 2022, Google announced a gradual transition to the Privacy Sandbox program on Android, aimed at ensuring maximum user data privacy while implementing marketing solutions. The Privacy Sandbox is still in the testing phase and hasn't been fully launched. So, what should marketers, ad networks, and MMPs expect, and how will this impact the market?
GAID (Google Advertising ID) is a unique identifier assigned to each mobile device, generated by Android to serve ads. While this identifier doesn’t contain personal user information, it does track user behaviour, transmitting this data to advertising networks. This level of anonymity ensures user privacy while maintaining the advertising ecosystem's functionality.
Google Advertising ID allows:
Refer to the table below for key insights.
| Purpose of Use | Who Uses It |
| Event-level reporting, aggregated reports | Marketers, MMPs |
| Remarketing | Ad nerworks, MMPs, marketers |
| Targeting | Ad networks |
From this table, we can conclude that the removal of Google Advertising ID will significantly impact marketing activities, especially in terms of remarketing and targeting. To better prepare for Google’s upcoming changes, let’s examine the importance of reporting, remarketing, and targeting within GAID.
GAID has been the primary source for analytical actions, but its deprecation does not mean that marketers and MMPs will lose access to reporting. So far, no announcements have been made about the discontinuation of the Google Play Install Referrer (GPIR). GPIR also upholds privacy standards and does not require sharing user identifiers with third parties. However, Google Play Install Referrer has the following limitations:
To prepare for the introduction of Privacy Sandbox, evaluating your current data setup is recommended, particularly identifying which installations currently rely on Google Play Install Referrer for attribution. This step can help alleviate concerns about major changes in attribution measurement.
Mobile remarketing involves displaying ads to users who have visited an app and performed key actions.
Remarketing heavily depends on Google Advertising ID (GAID), as the identifier allows advertisers to understand user interests and preferences, facilitating campaigns to re-engage former users.
User interactions with remarketing ads are shared with ad networks and publisher apps to optimize real-time algorithms. Further solutions from Google will be discussed later.
Third-party data exchanges via GAID help ad networks deliver ads tailored to each group’s interests, preferences, and behaviour. How does this work? For example, user 081's information is shared with the ad network. The data can then be used to target user 081 with an ad in another mobile app.
Currently, Google is aiming to reduce cross-platform targeting based on detailed user data. These new regulations could significantly impact ad networks once GAID is phased out.
Which ad networks you should use for in-app ads? Read here.
To safeguard user privacy, Google announced the release of Privacy Sandbox. Advertising networks often use trackers and other tools to gather information about user search queries, social media likes, and comments to create targeted banner ads.
The beta version of Privacy Sandbox offers API services that don’t rely on personal IDs but still track user activity across various apps. According to Google, Privacy Sandbox on Android will periodically assess user preferences based on installed apps. The software on mobile devices will be analyzed and categorized into areas like entertainment, finance, or movies. Based on this data, advertising priorities will be established.
With the discontinuation of GAID, what solutions does the Privacy Sandbox offer for analytics, retargeting, and targeting? Let's break it down.
To enable marketers and MMPs to obtain campaign performance data, Privacy Sandbox offers the Attribution API. Through this API, event-level reports and aggregated reports can be generated.
Event reports provide limited data at the time of app installation and afterwards. Unlike identifier-level reports, which offer specific signals for each ad, event-level reports have a delay of 1–2 days for view-through and click attribution, and 3 days for postbacks. These reports are more beneficial for ad networks than for marketers.
These reports provide more detailed user data, including the number of purchases, location, creatives, and ad names. Additionally, aggregated reports reflect the Lifetime Value (LTV) — the average revenue a user generates over the app's lifecycle. Postbacks are sent on the same day.
Despite the advantages of the Attribution API, it has some drawbacks:
As an alternative to GAID, Google suggests marketers use the Protected Audience API. This tool facilitates data sharing between the advertiser’s mobile app, publishers, and ad platforms. Audiences will be created on the device itself for personalized targeting, ensuring that third parties won’t have access to user data.
To prepare for these upcoming changes, we recommend focusing on the following:
To allow ad networks to continue delivering targeted ads, Google offers the Topics API. This feature enables a device to analyze user interactions with mobile apps over a specific period (currently 7 days) and select the top 5 topics of interest.
These topics will be determined based on the app’s name and description through a machine-learning classifier model, initially trained by Google.
The stricter privacy rules and upcoming changes with the Privacy Sandbox on Android are already stirring the mobile marketing industry. Marketers still recall the introduction of SKAdNetwork on iOS, which complicated user data collection in 2021.
To prepare for the discontinuation of Google Advertising ID, we recommend to start using aggregated data now.
