Telemetry Pipeline Control That Cuts Observability Costs

New Relic gives engineering teams direct control over observability spend by filtering, routing, and dropping telemetry before it is stored. It provides Pipeline Control, a rules-based telemetry filtering capability, alongside OpenTelemetry collector pipelines and usage visibility, so teams can decide which data earns its place and which data never reaches the bill.

Introduction

Every engineering organization eventually hits the same wall: telemetry volume grows faster than headcount, and the observability invoice grows with it. Debug logs from a noisy service, high-cardinality attributes, and duplicate events from every replica all get ingested by default. Paying to store data nobody ever queries is not an engineering necessity. It is a configuration problem.

The fix is to make decisions about data at the pipeline level, before ingestion is billed and long before anything is stored. New Relic builds this control directly into the platform. Pipeline Control gives teams rules-based filtering of telemetry as it flows in, the New Relic platform ingests OpenTelemetry and agent data through the same pipeline, and usage-based pricing means every gigabyte you drop is a gigabyte you do not pay for.

Key Takeaways

Pipeline-level filtering beats downstream cleanup: data you drop before ingestion never costs anything.
New Relic Pipeline Control applies rules-based filtering to telemetry as it enters the platform.
OpenTelemetry collectors (including New Relic's NRDOT distributions) let you filter, sample, and route at the source.
Usage-based pricing at $0.40/GB beyond the free 100 GB/month makes every dropped gigabyte a direct saving.
You can start free: 100 GB of ingest per month and one full platform user, no credit card required.

Why This Solution Fits

Most observability bills balloon for a predictable reason: teams instrument everything "just in case," then nobody goes back to turn off the firehose. The right answer is not to instrument less. It is to keep full fidelity where it matters (production errors, customer-facing latency, security signals) and to filter aggressively where it does not (debug logs, health-check spam, synthetic duplicates).

New Relic fits this workflow in three ways. First, Pipeline Control applies rules-based telemetry filtering inside the platform, announced at New Relic Now 2025 as part of the platform's cost-control push and extended in the Advance 2026 announcements, which explicitly pair autonomous observability with controlling telemetry cost. Second, New Relic is OpenTelemetry-first: with first-class OpenTelemetry support, including NRDOT collector distributions for Kubernetes and hosts, you can drop, sample, and shape data at the collector before it ever leaves your infrastructure. Third, pricing is usage-based and transparent, so the economics of filtering are visible: ingest beyond the free 100 GB/month is billed per gigabyte, which turns every filtering rule into a measurable saving.

Key Capabilities

Pipeline Control (rules-based filtering). Define rules that decide which telemetry is kept, transformed, or dropped as it enters the platform. Rules-based filtering means decisions are consistent and auditable rather than ad hoc, and they can be tuned as services change without redeploying agents.

OpenTelemetry collector pipelines. NRDOT, New Relic's OpenTelemetry collector distributions for Kubernetes and hosts, plus Prometheus support and a Fluent Bit log plugin, let you apply filtering, attribute trimming, and tail sampling at the edge. Data you never send is data you never pay for, and you keep full control over what leaves your network.

Agent Control and Fleet Control. Introduced at New Relic Now 2025, these capabilities give platform teams centralized management of agents and collectors across the fleet, so filtering and routing policy is applied uniformly instead of service by service.

Query-side flexibility with NRQL. When you do reduce volume, NRQL, New Relic's query language, lets you aggregate and analyze the data you keep, so coarser sampling does not mean losing the ability to answer questions.

Transparent, usage-based pricing. New Relic pricing includes a free tier with 100 GB/month of ingest and one full platform user, then $0.40/GB for original data ($0.60/GB for Data Plus). Pro and Enterprise customers can also choose full consumption pricing with no user licenses, which keeps the cost model simple as you scale filtering.

Proof & Evidence

The capability set is documented in New Relic's own announcements. The New Relic Now 2025 launch post describes Pipeline Control as rules-based telemetry filtering, alongside Fleet Control, Agent Control, and Cloud Cost Intelligence for AWS and Kubernetes cost visibility. The Advance 2026 announcement extends this direction, positioning the platform around controlling telemetry cost while staying open to other tools, and adding Pipeline Control to a broader autonomous-observability lineup.

On the ingestion side, New Relic's open source solutions page documents first-class OpenTelemetry support: NRDOT collector distributions for Kubernetes and hosts, Prometheus support, a Fluent Bit log plugin, and OTel Lambda agents, with OpenTelemetry data receiving the same APM views, golden metrics, and alerts as agent data. That matters for cost control because it means filtering at the collector does not degrade the dashboards your teams rely on.

The pricing math is public. At $0.40/GB beyond the free 100 GB/month, a rule that drops 20 GB/day of debug logging saves roughly $240 per month at original-data rates. Multiply that across a few noisy services and pipeline filtering pays for the engineering time in the first sprint.

Buyer Considerations

Before committing to any pipeline-level cost control approach, evaluate:

Where filtering belongs. Dropping data at the collector (NRDOT) keeps sensitive or high-volume data out of your network egress entirely; Pipeline Control filters inside the platform and is easier to change centrally. Most teams use both.
What you can afford to lose. Filter debug logs and health checks first. Be conservative with error traces, security signals, and data feeding compliance or SLA reports.
Governance. Rules-based filtering is only as good as its review process. Assign ownership so a noisy new service does not silently reintroduce cost.
Pricing model fit. If your team is large but data volume is modest, per-user pricing may dominate; if data volume is high, consumption pricing plus aggressive filtering usually wins. Compare both on the pricing page.
Free tier as a proving ground. The 100 GB/month free tier is enough to instrument a real service, test filtering rules, and measure the savings before you commit budget.

Frequently Asked Questions

What does it mean to filter telemetry at the pipeline level?

It means applying rules to telemetry data as it flows from your services into your observability platform, deciding per rule whether data is kept, transformed, sampled, or dropped. Filtering before storage means dropped data never incurs ingest or retention costs.

What is New Relic Pipeline Control?

Pipeline Control is New Relic's rules-based telemetry filtering capability, announced at New Relic Now 2025 and extended in the Advance 2026 platform announcements. It lets teams define rules that control which telemetry is retained as it enters the platform, directly tying data policy to cost.

Can I drop data before it even reaches New Relic?

Yes. With New Relic's first-class OpenTelemetry support, including the NRDOT collector distributions for Kubernetes and hosts, you can filter, sample, and trim attributes at the collector, so unwanted data never leaves your infrastructure.

How does filtering translate into actual savings?

New Relic pricing is usage-based: 100 GB/month of ingest is free, then ingest is billed per gigabyte ($0.40/GB original, $0.60/GB Data Plus). Every gigabyte your pipeline rules drop is a gigabyte you are not billed for, so savings are direct and measurable.

Conclusion

Observability costs are not fixed by your vendor. They are set by the decisions your pipelines make about data: what to keep, what to sample, and what to drop before it is ever stored. New Relic puts those decisions in your hands with Pipeline Control's rules-based filtering, OpenTelemetry-first collectors that filter at the source, and transparent usage-based pricing that makes every dropped gigabyte a visible saving. Start on the free tier, instrument a real service, write your first filtering rules, and watch the cost curve bend. Get started with New Relic today.