What Privacy Sandbox Is
Privacy Sandbox is Google's multi-year initiative to develop privacy-preserving technologies that replace the tracking mechanisms the digital advertising industry has relied on for decades. On the web, it replaces third-party cookies. On Android, it replaces the Google Advertising ID. The core premise is that advertising measurement, targeting, and fraud prevention can work effectively without giving advertisers access to individual user identifiers.
The initiative spans both Chrome (web) and Android (mobile), though the implementations differ significantly. On the web side, Privacy Sandbox has been in development since 2019 and includes proposals like Topics API, Attribution Reporting API, Protected Audiences API (formerly FLEDGE), and several others. On Android, the Privacy Sandbox was announced in early 2022 and introduces mobile-specific versions of these APIs adapted for the app ecosystem.
For mobile growth teams, the Android Privacy Sandbox is the critical piece. It represents the most significant change to Android advertising infrastructure since GAID was introduced. Unlike Apple's approach with ATT, which was a relatively sudden shift, Google is pursuing a more gradual transition, maintaining GAID availability while the Privacy Sandbox APIs mature and the ecosystem adapts. This gives teams more time to prepare, but it also means the transition period will be longer and more complex to manage.
Key Privacy Sandbox APIs for Mobile
The Attribution Reporting API is the most directly relevant API for growth teams focused on campaign measurement. It provides two types of reports: event-level reports that link a specific ad interaction to a conversion with limited conversion data, and aggregate reports that provide detailed conversion data without linking it to individual users. Event-level reports are useful for optimization signals, while aggregate reports support more detailed analysis of campaign performance across dimensions like geography, creative, and audience segment.
The Topics API replaces interest-based targeting that previously relied on cross-app tracking. Instead of building detailed user profiles from browsing and app usage history, Topics assigns a small number of interest categories to each user based on their recent app usage. Advertisers can target ads based on these topics without learning anything else about the user. The topics are coarse-grained by design, categories like "Sports" or "Travel" rather than specific behaviors, which limits targeting precision but preserves user privacy.
Protected Audiences API (formerly FLEDGE) enables remarketing and custom audience targeting without sharing user data with third parties. The audience membership and ad selection logic runs on-device rather than on remote servers. This means an advertiser can show ads to users who previously visited their app without that user's identity or behavior being exposed to the ad network or any intermediary. The on-device auction model is a fundamental architectural shift from how remarketing has traditionally worked.
How Privacy Sandbox Changes Attribution
The shift from GAID-based deterministic attribution to Privacy Sandbox attribution reports requires growth teams to rethink how they measure and optimize campaigns. The most significant change is the move from user-level to aggregate measurement as the default. While GAID allowed you to trace the complete journey of an individual user from ad click through install to every subsequent in-app event, Privacy Sandbox attribution reports provide conversion data at a campaign or ad group level with intentional noise added to prevent re-identification.
Event-level reports from the Attribution Reporting API include the ad interaction source (click or view) and a limited conversion value, but they are subject to delays and noise. You cannot reconstruct a complete user journey from these reports. Aggregate reports provide richer conversion data, revenue values, event counts, conversion paths, but only in aggregate form that cannot be tied back to individual users. The reports use cryptographic techniques to ensure that the aggregated data is accurate in total while preventing extraction of individual records.
Linkrunner is built to work seamlessly with Privacy Sandbox attribution signals alongside traditional GAID-based attribution and other measurement methods. As the Android ecosystem transitions from GAID to Privacy Sandbox, teams using Linkrunner maintain consistent campaign measurement without needing to rebuild their analytics infrastructure for each new API. The platform translates between attribution methodologies, giving growth teams a stable measurement layer even as the underlying signals evolve.
Privacy Sandbox vs. Apple's Approach
Comparing Google's Privacy Sandbox with Apple's ATT and SKAdNetwork reveals fundamentally different philosophies toward the same goal. Apple chose a consent-based approach: users explicitly opt in or out of tracking, and the alternative measurement system (SKAdNetwork) provides minimal data with strict privacy guarantees. Google chose a technology-based approach: replace the tracking mechanism itself with new APIs that provide useful advertising signals without exposing individual identifiers.
In practice, Privacy Sandbox offers more granular data than SKAdNetwork. SKAdNetwork provides a single conversion value per install with significant time delays and no view-through attribution in earlier versions. Privacy Sandbox's Attribution Reporting API supports both click-through and view-through attribution, provides multiple conversion reports per user journey, and offers both event-level and aggregate reporting options. This gives growth teams more data to work with for optimization.
The transition timelines also differ dramatically. Apple's ATT was enforced within months of announcement, giving the industry limited time to adapt. Google has taken a multi-year approach, running Privacy Sandbox APIs alongside GAID and committing to advance notice before any GAID deprecation. This gradual approach reduces disruption but also creates a prolonged period of uncertainty where teams must support both old and new measurement systems simultaneously.
For growth teams running cross-platform campaigns, these differences mean that iOS and Android measurement will continue to operate under different paradigms for the foreseeable future. Building a measurement framework that normalizes data across both platforms, accounting for the different signal types, latency characteristics, and granularity levels, is essential for making accurate cross-platform comparisons and allocation decisions.
Preparing Your Measurement Stack
Preparing for Privacy Sandbox is not a single migration, it is an ongoing process of adapting your measurement infrastructure to work with evolving APIs. Start with an audit of your current GAID dependencies. Map every system that reads, stores, or processes GAID data: attribution platforms, analytics tools, audience builders, fraud detection systems, and data warehouses. For each dependency, identify the Privacy Sandbox API that will serve as the replacement.
Test the Attribution Reporting API in your development environment now. Register your app as an attribution source and destination, implement the API calls for triggering attribution, and examine the reports you receive. Understanding the data format, latency, and noise characteristics of these reports before you depend on them in production is critical. The reports look and behave differently from GAID-based attribution data, and your analytics pipelines will need adjustments.
Invest in server-side infrastructure for aggregate report processing. Privacy Sandbox aggregate reports require an aggregation service that runs in a trusted execution environment. This is a new infrastructure component that most growth teams have not needed before. Work with your attribution provider to understand how they handle aggregate report processing, or evaluate whether you need to build this capability in-house for custom analysis.
Finally, update your team's mental models. Privacy Sandbox is not just a technical migration, it requires a shift in how growth teams think about measurement. The era of tracing individual user journeys from click to lifetime value is ending on Android, just as it ended on iOS with ATT. Teams that embrace aggregate measurement, invest in statistical modeling, and supplement attribution data with incrementality testing will maintain their competitive edge through this transition.
