Product Analytics
Product analytics is the process of analyzing how users engage with a digital product. While traditional System Observability (Prometheus, Loki) focuses on “Is the system up?”, Product Analytics focuses on “What are users doing and why?“.
Key Concepts
1. The “Whose” vs. “What”
- Infrastructure Monitoring: Tracks CPU, memory, and error rates (The “What”).
- Product Analytics: Tracks user identity, journey paths, and feature adoption (The “Whose”).
2. Events & Properties
- Events: Actions taken by a user (e.g.,
Page Viewed,Build Started,Search Performed). - Properties: Metadata attached to events (e.g.,
device_type,build_duration,file_count).
3. Autocapture vs. Instrumentation
- Autocapture: Automatically recording every click and interaction (e.g., PostHog).
- Manual Instrumentation: Specifically writing code to track high-value business actions.
Why it matters for DevSecOps
For an engineer, product analytics bridges the gap between Code and User Value. By tracking how often a CLI tool is used or where a build fails, we can prioritize infrastructure hardening where it actually impacts the user.
Product Analytics as a Pillar of Observability
Modern Observability is evolving beyond the “Three Pillars” (Metrics, Logs, Traces). By including Product Analytics as a fourth pillar, we achieve Full-Stack Visibility:
- System Health: “Is the Kubernetes cluster healthy?” (Prometheus).
- Code Health: “Are there exceptions in the application logs?” (Loki).
- User Health: “Are users actually able to complete their intended workflow?” (PostHog).
When a system is “green” in Grafana but users are dropping off in PostHog, it signals a functional failure—an observability gap that infrastructure metrics alone cannot see. Adding user behavior to the observability stack ensures we are monitoring the outcome, not just the output.
Related Tools
- PostHog
- Plausible
- Google Analytics
- Mixpanel