Slide 1 of 2

Reduce costs with Adaptive Telemetry

Optimize your data collection

Observability costs tend to grow with the data you collect, and that growth can be hard to predict. You can reduce the telemetry you pay for but don’t use with Adaptive Telemetry, a service available only in Grafana Cloud.

It uses machine learning to identify data that isn’t referenced in dashboards, alerts, or queries, and recommends ways to reduce it. You review and apply the recommendations:

  • Nothing is dropped automatically.
  • You can set exemptions to protect any data you want to keep.

How Adaptive Telemetry reduces costs

Each tool targets a different signal type but follows the same approach: find what’s unused and recommend a reduction.

  • Adaptive Metrics: Finds metrics with many unique label combinations (high cardinality) that can be aggregated or dropped
  • Adaptive Logs: Finds high-volume, low-value log patterns that can be filtered before they reach Grafana Cloud storage
  • Adaptive Traces: Samples traces intelligently, keeping errors and high-latency events while dropping repetitive healthy traces

Real results from real teams

  • Property Finder

    “Adaptive Telemetry helped us optimize costs without constantly waiting for engineers to fix instrumentation… Adaptive Tracing allowed us to shift trace sampling policies left. That allowed us to control costs while keeping engineers fully in the loop.”

  • Lovable

    “Adaptive Metrics helped us drastically reduce costs while keeping the visibility we needed.”

Script

Observability costs grow with the data you collect, and that growth is hard to predict. Adaptive Telemetry uses machine learning to find data that isn’t referenced in any dashboard, alert, or query, then recommends ways to reduce it.