Diagnosing Android FCM Delivery Latency & Doze Mode Throttling

When mobile backend teams send transactional push alerts via Firebase Cloud Messaging (FCM), a 200 OK response from the Google FCM endpoint merely confirms that Google accepted the message into its dispatch queue. It does not indicate that the target Android device has received or rendered the alert.

In high-urgency domains—such as delivery dispatch, ride-hailing, real-time security approvals, and financial margin calls—push notifications frequently experience latency delays ranging from 30 seconds to over 15 minutes.

How Android Doze Mode Intercepts Push Messages

Introduced to preserve battery life, Android Doze mode puts idle devices into deep sleep states when unplugged and stationary. During deep Doze, the operating system suspends network access, defers background jobs, and blocks standard broadcast intents.

FCM provides two message priority tiers that interact differently with Doze:

1. Normal Priority (priority: "NORMAL"): FCM attempts immediate delivery if the device is active. If the device is in Doze, FCM holds the message and batches delivery during the next periodic Doze maintenance window.
2. High Priority (priority: "HIGH"): FCM wakes the device radio, bypasses Doze restrictions, and invokes onMessageReceived() immediately. However, Google enforces a rolling quota on high-priority messages; if an application dispatches excessive high-priority alerts without prompting user interactions, FCM downgrades subsequent messages to normal priority.

OEM-Specific Battery Management Profiles

Beyond stock Android Doze, major hardware manufacturers implement proprietary battery optimization daemons:

• Xiaomi (HyperOS / MIUI): Aggressively halts background service processes and restricts background autostart unless explicitly granted by the user.
• Samsung (OneUI): Places infrequently opened applications into “Deep Sleeping Apps” status, disabling FCM wakeups entirely until the application is manually launched.
• Huawei / Honor: Enforces custom power management whitelists that block background broadcast receivers unless battery optimization is disabled.

Recommended Gateway & Client Architecture

To achieve sub-second delivery latencies on Android devices:

• Set FCM Priority Explicitly: Ensure all real-time transactional payloads declare "android": { "priority": "high" } in the FCM v1 JSON payload.
• Include Direct Boot Compatibility: Mark your FirebaseMessagingService with android:directBootAware="true" in AndroidManifest.xml to receive messages before the device is unlocked following a reboot.
• Instrument Display Timestamps: Measure the delta between server dispatch timestamp t_dispatch and device receipt timestamp t_received inside onMessageReceived() to log real-world P99 latency percentiles.

If your organization requires a multi-device latency diagnostic, review our Delivery Latency service or schedule a consultation.