Maximize data value and cut costs: Adaptive Telemetry for metrics, logs, traces, and profiles in Grafana Cloud

When it comes to observability, more data doesn’t always mean more clarity. In fact, as telemetry volumes grow, it only becomes more difficult to discern the signals from the noise and to keep overall costs in check.

This is exactly why we built Adaptive Telemetry, a suite of features in Grafana Cloud that analyzes how your telemetry is used and then automatically recommends actions like aggregating, sampling, dropping, or reducing low-value data. You can easily pinpoint the critical signals you need to keep your systems running smoothly, while reducing observability costs.

Our Adaptive Telemetry suite started with Adaptive Metrics and Adaptive Logs — and our users have seen some pretty incredible results from both of these solutions, from improved productivity to reducing their observability costs by 50% (without losing visibility).

Now, at ObservabilityCON 2025, we’re excited to share that our Adaptive Telemetry suite is evolving, with Adaptive Traces now generally available and Adaptive Profiles in private preview.

With this end-to-end Adaptive Telemetry suite, you can extend the benefits of the adaptive model across all core pillars of your observability strategy, ensuring every piece of telemetry you store is truly worth the investment.

Adaptive Traces: Retain only your most valuable traces 

In modern distributed systems, tracing provides visibility into how requests move across services, exposing bottlenecks and interdependencies. But once tracing expands beyond a few services, costs rise, dashboards grow noisy, and teams often abandon the practice altogether — not due to lack of value, but because successful requests generate large volumes of redundant data that’s expensive to store.

Fortunately, there’s a better way. With Adaptive Traces in Grafana Cloud, you can ensure that only valuable traces, such as those with errors, high latency, or other critical signals, are stored, leading to faster insights and optimized costs.

“Before Adaptive Tracing, we had two bad options: send everything and blow our budget, or send so little we couldn’t get meaningful insight,” said Geoff Schultz, Manager, Infrastructure Engineering at Auditboard. “Now tracing is actually usable — we can dial sampling up or down as needed, keep costs in check, and still give teams the visibility they need.”

Adaptive Traces achieves this using tail sampling, a technique where the decision to sample (aka, keep) or drop a trace is made after collecting all or most of its spans, rather than at the start of the trace. This allows for light sampling during normal operations, and then more targeted retention when issues arise.

Custom policy management for fine-grained control

Adaptive Traces offers out-of-the-box recommendations, while also allowing you to define flexible and customizable sampling policies to capture only the tracing data you need. For example, you might configure rules to ingest all traces with an error status from a particular service, or to capture a random 5% of traces from another.

By automatically filtering out noise while retaining the traces that matter most, Adaptive Traces allows you to debug more effectively and get deeper, more meaningful insights into system performance.

Anomaly detection to capture traces of interest 

Adaptive Traces’ anomaly detection feature automatically identifies anomalies in your tracing data, and then adjusts sampling decisions and policies on-the-fly, based on those observed anomalies.

To do this, Adaptive Traces analyzes your past data to establish a baseline for “normal” system behavior, such as the typical latency for a specific database query. It then monitors incoming traces for significant changes, like a latency spike, from that established baseline.

When a trace is identified as an anomaly, it’s retained rather than dropped, helping you more easily identify issues.

Closed-loop investigations with Traces Drilldown 

Adaptive Traces now integrates directly with Grafana Traces Drilldown, an application within our broader Grafana Drilldown suite that allows you to quickly investigate and visualize your tracing data through a simplified, queryless experience.

When an anomaly is detected, click Drilldown within the anomaly policy view to instantly jump to relevant traces within the time range the issue occurred. This provides a closed-loop investigative workflow that allows you to quickly find root causes and reduce time to recovery.

Adaptive Profiles: The power of continuous profiling without runaway costs

Continuous profiling is a critical piece of any observability strategy, providing deep insights into how your application uses resources, like CPU and memory, in production. Because data is collected continuously, you get a holistic and uninterrupted view of your application’s behavior over time.

And now, with Adaptive Profiles, teams can deploy continuous profiling more broadly across their infrastructure, maximizing the value of their data without the risk of cost overruns.

Adaptive Profiles dynamically adjusts the detail and frequency of data collection based on workload behavior, then generates insights to help your developers optimize their services’ performance and cost. It ensures you capture critical, high-resolution data to generate performance insights when you need them, and then reduces data to a cost-effective baseline during normal operations.

To try Adaptive Profiles in Grafana Cloud, please fill out this private preview interest form.

Rounding out the suite: Adaptive Metrics and Adaptive Logs

Adaptive Traces and Adaptive Profiles can work alongside two other powerful features in Grafana Cloud — Adaptive Metrics and Adaptive Logs — to help you filter out noise and reduce observability costs.

Adaptive Metrics continuously analyzes how your metrics are being used across dashboards, alerts, recording rules, and queries, and provides recommendations for aggregating underutilized metrics into lower-cardinality versions.

Adaptive Logs, meanwhile, is a cost optimization feature that analyzes log data at scale, identifies commonly ingested log patterns, and generates customized recommendations for dropping unused or low-value logs.

You can learn more about both features, and how they’ve benefitted our users, in this blog post.

Get started with Adaptive Telemetry today

You can easily get started today with Adaptive Traces, Adaptive Logs, and Adaptive Metrics in all tiers of Grafana Cloud, including in our free tier. Simply navigate to Adaptive Telemetry within the Grafana Cloud UI.

If you’re interested in the Adaptive Profiles private preview, please complete this form and we’ll be in touch.

To read all the news from ObservabilityCON 2025, check out our blog post recapping all the big announcements!

FAQ: Adaptive Telemetry in Grafana Cloud

What is Adaptive Telemetry?

Adaptive Telemetry is a suite of features in Grafana Cloud designed to optimize observability data by ensuring that only the most valuable telemetry — such as metrics, logs, traces, and profiles — is stored and surfaced. It uses intelligent classification and prioritization techniques to analyze how telemetry is used (e.g., in dashboards, alerts, or queries) and then automatically recommends actions like aggregating, sampling, dropping, or reducing low-value data.

This approach helps organizations filter out noise, reduce costs, and maintain the insights needed for effective monitoring and troubleshooting.

What is included in the Adaptive Telemetry suite?

The Adaptive Telemetry suite in Grafana Cloud consists of four features, spanning all core pillars of observability:

• Adaptive Metrics: Helps you identify and eliminate unused or partially used time series data through aggregation.
• Adaptive Logs: Reduces log volume and associated costs by automatically identifying and removing low-value logs that are rarely or never used.
• Adaptive Traces: Uses tail sampling to automatically identify and retain your most valuable traces, ensuring you get the critical performance insights you need.
• Adaptive Profiles: Dynamically adjusts data collection based on workload behavior, allowing you to deploy continuous profiling more broadly across your infrastructure without incurring excessive costs.

What are the key benefits of Adaptive Telemetry?

Adaptive Telemetry is especially beneficial for high-cardinality and dynamic environments, helping teams maintain visibility while controlling costs. Here are some of the highlights:

• Keeping only what’s valuable: Adaptive Telemetry classifies telemetry based on actual usage, system behavior, and human inputs, prioritizing signals that power alerts, dashboards, or are frequently queried, while identifying and optimizing away less useful data.
• Cost and noise reduction: By focusing on valuable telemetry, Adaptive Telemetry reduces alert fatigue and significantly lowers storage and processing costs. For example, on average, organizations using Adaptive Metrics see a 30% - 50% reduction in telemetry costs.
• Automated and flexible optimization: The system continuously learns from usage patterns and adapts its recommendations. You can review and implement optimizations in minutes based on your requirements.
• Fine-grained control: With features like exemptions in Adaptive Metrics, you can be sure critical data and signals are always retained.

Where can I learn more about Adaptive Telemetry?

You can learn more about Adaptive Telemetry in our documentation, this blog post, and this overview video.

Grafana Cloud is the easiest way to get started with metrics, logs, traces, dashboards, and more. We have a generous forever-free tier and plans for every use case. Sign up for free now!