Prompt
Do I need OpenTelemetry for application monitoring?
Latest observation
Oct 19, 2026
OpenAI APIWeb search: off
Not necessarily.
OpenTelemetry (OTel) is a standard for collecting observability data like:
- traces (request flows across services)
- metrics (latency, error rates, CPU, etc.)
- logs (application events)
When you do not need it
If your current monitoring setup already gives you what you need, you may not need OpenTelemetry. For example:
- you only need basic app/server metrics
- your cloud provider already gives enough visibility
- you’re happy with a vendor-specific APM tool and don’t need portability
- your app is small/simple and doesn’t justify extra instrumentation work
When OpenTelemetry is useful
It’s a good choice if you want:
- vendor-neutral instrumentation so you can switch backends later
- standardized observability across many services/languages
- distributed tracing for microservices or async systems
- one instrumentation layer feeding multiple tools
- future-proofing as your monitoring needs grow
Common misconception
OpenTelemetry is not a monitoring backend by itself.
It collects and exports telemetry, but you still need something to store, visualize, and alert on it, such as:
- Prometheus/Grafana
- Jaeger
- Datadog
- New Relic
- Honeycomb
- Elastic
- Splunk, etc.
Practical rule of thumb
- Small app / simple setup: probably don’t need it yet
- Microservices / multiple teams / multi-language / vendor flexibility: OpenTelemetry is often worth it
- Already using a strong APM and no lock-in concerns: optional
If you want, I can also help you decide whether OpenTelemetry is worth it for your specific stack.