Privacy-First Telemetry: Tracking Multi-Device Usage Without Leaking PII

The Modern Privacy Landscape

The era of tracking users across devices using persistent hardware serials (IDFA, GAID, IMEI) or passing clear-text email addresses in telemetry payloads is definitively over.

Under Thailand’s Personal Data Protection Act (PDPA), the EU’s GDPR, and Apple’s App Tracking Transparency framework, telemetry pipelines must be engineered with privacy by design. Organizations that fail to decouple personal identifiers from behavioral telemetry risk substantial regulatory penalties and App Store rejections.

However, strict privacy compliance does not require sacrificing accurate cross-platform usage reporting.

The Pseudonymized Token Hierarchy

To maintain user cohort visibility without storing Personally Identifiable Information (PII) in your analytical data warehouse, establish a three-tier token hierarchy:

[User PII (Auth DB Only)] 
          |
          v  (One-way HMAC-SHA256 with Salt Rotation)
[Hashed Analytics User ID]
          |
          +---> [Session UUID (Rotates every 15m of inactivity)]
          +---> [Device Instance UUID (Vendor Identifier, Reset on Reinstall)]

1. The Salted Hash Analytics Identifier

Never pass database primary keys or email addresses as the user_id attribute. Instead, generate a one-way HMAC-SHA256 hash using a server-side pepper stored exclusively in your authentication backend.

2. Ephemeral Session Tokens

Session IDs must be randomly generated UUIDs that rotate whenever a session boundary is crossed. This prevents unauthorized third parties from correlating unrelated user sessions over extended periods.

3. Granular Client-Side Scrubbing

Implement automated property filters in your client telemetry wrappers that scan event payloads and strip out forbidden fields (such as phone numbers, credit card strings, or precise GPS coordinates) before network transmission.

Zero-PII Telemetry Payload Sample

{
  "timestamp": "2026-08-11T04:12:30.120Z",
  "client_platform": "ios",
  "app_build": "3.4.1",
  "os_version": "19.5",
  "analytics_id": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
  "session_id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
  "event_name": "Course Module Completed",
  "properties": {
    "module_category": "engineering",
    "duration_seconds": 340,
    "completed_via_offline_cache": false
  }
}

Compliance and Precision Can Coexist

By decoupling identity from behavior at the SDK level, engineering teams achieve the best of both worlds: full legal compliance with international privacy mandates, and mathematically robust usage metrics that accurately reflect real-world product adoption.