Hey everyone! We recently put together a blog post on how engineers can go beyond the basics with OpenTelemetry by using custom metrics to get deeper insights into application behaviour. It covers practical examples like tracking user drop-offs, understanding feature usage, and tying metrics to actual business outcomes (like reducing cart abandonment). All based on real use cases that is worked through using OpenTelemetry and SigNoz. If you're exploring how to make your observability setup more actionable, especially beyond infra/ default metrics, this might be helpful. Happy to hear thoughts or experiences from others doing similar things. Check the blog out! 👀

IMHO the missing part of the opentelemetry protocol is authorisation of the spans. Why would you mix the spans that can be manipulated from the frontend side, and the spans from your internal infrastructure. I think the protocol or the opentelemetry implementations should add a way to authenticate the comming spans by some token, so you can be sure that somebody sending bogus spans from your instrumented frontend app does not endup in your internal infrastructure dashboards.