IAB Tech Lab has unveiled a significant enhancement to its Open Measurement SDK for Android with the release of version 1.6.10. This update introduces a new capability that allows verification scripts for native ads to operate within Android’s Jetpack JavaScriptEngine, a sandboxed environment, rather than the traditional WebView. This change, articulated in a technical post by Sid Polo, a senior software engineer at DeliveryHero, is poised to streamline ad verification processes, particularly on lower-end devices.
In Short
As of October 1, 2026, Android applications can now utilize a new option that enables them to verify ad visibility through a more efficient sandboxed engine. This transition has demonstrated a remarkable reduction in the time taken to initiate verification checks—approximately 11 milliseconds in the sandbox environment compared to 268 milliseconds in the WebView. This improvement is crucial for app publishers, as it allows for smoother ad loading and measurement during critical moments of app performance. Notably, applications that do not adopt this new feature will continue to function as they always have, and HTML-based ads remain outside the current scope of this update.
What the post describes
The historical context of the Open Measurement SDK indicates that verification scripts for native ad sessions were previously confined to the WebView. However, with the introduction of version 1.6.10, these sessions can now leverage the androidx.javascriptengine, which is recommended by Android for non-interactive JavaScript execution outside of a WebView. The post emphasizes that this new method is “opt-in,” faster to initiate, and less likely to disrupt app performance.
Two design features ensure that apps opting out of this change are minimally affected. The first is fallback: if the sandbox is unavailable, the SDK will revert to the existing WebView executor. The second is scope: apps that do not integrate the sandbox will operate as they did previously, ensuring a non-disruptive transition. The team is also exploring future support for HTML and JavaScript ad sessions.
The benchmark figures
A benchmark test conducted on a low-end device revealed striking differences in performance across three scenarios. The baseline run, which did not create any OM session, served as a control to measure the overhead of session creation.
| Run | First session creation (median) | Frame overrun (P99) |
|---|---|---|
| Baseline (no OM session) | 0 ms | 114.51 ms |
| JSEngine | 11.07 ms | 119.08 ms |
| WebView | 268.51 ms | 589.07 ms |
The data indicates that the median first-session creation time on the WebView path was 24 times longer than that of the JavaScriptEngine path, highlighting a significant efficiency gain. For app publishers, these improvements translate into smoother ad-load experiences, particularly on lower-end devices where the cost of initializing a cold WebView can be substantial.
How Android’s sandbox works
According to Android’s developer documentation, the JavaScriptEngine library facilitates JavaScript evaluation without necessitating a WebView instance. This method offers several advantages, including reduced resource consumption, compatibility with background services, and the ability to maintain multiple isolated environments with minimal overhead.
Sandbox and isolates
The design is anchored by two key components: a JavaScriptSandbox, which connects to the out-of-process engine, and a JavaScriptIsolate, which serves as the execution context within the sandbox. Applications must first verify sandbox support before proceeding with any operations, and creating multiple sandbox instances is restricted to one per application.
Crash behaviour
All JavaScript executions occur within a separate sandboxed process. This isolation ensures that if a crash occurs within the sandbox, it does not impact the main application process. The post outlines various exceptions that may arise from sandbox crashes, emphasizing the importance of robust error handling.
Optional features and data limits
The documentation notes that the availability of certain features may vary based on the WebView version, necessitating checks for feature support. Data handling capabilities are also governed by specific features, with guidelines on how to manage large data transfers effectively.
Adoption as the post lays it out
The adoption process for this new feature involves four key steps: updating to OM SDK Android 1.6.10 or later, integrating the androidx.javascriptengine dependency, establishing a connected JavaScriptSandbox on API 26 or higher, and maintaining the sandbox throughout the app’s lifecycle. The post provides a code snippet for activation, illustrating the straightforward implementation process.
Where the change sits in the OM SDK’s history
The Open Measurement SDK has evolved significantly since its initial release in April 2018, achieving widespread adoption across billions of devices. Subsequent updates have introduced various capabilities, including audio ad measurement and device attestation. The latest enhancements reflect ongoing efforts to improve measurement accuracy and efficiency within the mobile advertising ecosystem.
Why the executor matters beyond developers
For publishers and agencies involved in app inventory purchasing, the implications of this change extend beyond technical specifications. The performance of verification sessions directly influences the ad rendering experience, with the potential for varying execution environments across different devices and applications. The post leaves open questions regarding the extent of adoption and the impact on overall ad performance metrics.
