Documentation Grafana Cloud Cost management and billing Adaptive Telemetry Adaptive Traces Guides Focus on value, not volume

Best practice - focus on value, not volume

It can be tempting to measure the success of Adaptive Traces by looking only at the drop rate metric.

A common misconception is that a high drop rate is the only indicator of significant cost savings. However, this approach overlooks the primary value of the feature: intelligently preserving the most critical troubleshooting data.

Tail-based sampling ensures that you are capturing all critical traces. It analyzes a complete trace for important signals—like errors or high latency—before making any keep/drop decision. Based on your configured policies, 100% of traces that match your critical criteria are automatically preserved in their entirety, ensuring you always have the full context you need for any investigation.

The true cost savings come from not paying to store and index low-value data, while the true value comes from having every critical trace you need at your fingertips during an investigation.

Instead of focusing solely on the volume of dropped data, best practice is to evaluate success based on these outcomes:

• Are you monitoring the effectiveness of your Adaptive Traces configuration using the built-in metrics? They help you verify that critical signals are being preserved while low-value data is being dropped.
• Are you capturing all critical traces? Verify that you have visibility into traces with errors or high latency.
• Is your Mean Time to Resolution (MTTR) decreasing? Are your teams solving problems faster because they have exactly the data they need, without having to sift through the noise of millions of redundant traces?