OpenTelemetry (OTel) is the standard for collecting telemetry data, but you still need a backend to store, query, and visualize it. We compare the top OTel-compatible backends — Grafana LGTM, Datadog, Honeycomb, and New Relic — across cost, scalability, and team expertise so you can pick the right one for production.
OpenTelemetry (OTel) has become the de facto standard for collecting traces, metrics, and logs from your applications. It handles the plumbing — instrumentation, sampling, export — but it doesn't store or analyze anything. For that, you need a backend.
The market is crowded, and choosing the wrong one means either overpaying for unused capacity or hitting a wall when your data volume grows.2 Here's how the top OTel backends stack up for production workloads.
Grafana's LGTM stack — Loki (logs), Grafana (visualization), Tempo (traces), Mimir (metrics) — is the most complete open-source OTel backend available. All four components accept OTLP natively, so you can send everything through a single pipeline.1
Self-hosting LGTM gives you full control over retention, sampling, and cost. The trade-off is operational complexity: you're running four stateful services, plus object storage (S3/GCS) underneath. Grafana Cloud offers a managed version if you want the same stack without the ops burden.
Best for: teams that want open-source flexibility, have DevOps chops, and need to keep costs predictable at scale.
Datadog ingests OTel data through its OTLP intake endpoint and maps it into its own data model, giving you access to its full suite of dashboards, monitors, and APM features.3 Distributed tracing is particularly strong — Datadog can correlate a single trace across services, infrastructure, and logs without manual configuration.
The downside is pricing. Datadog charges per host and per ingested volume, and costs can escalate quickly as you add services or increase sampling rates. It's the right choice when you need enterprise support, compliance certifications, and a single pane of glass across the entire stack.
Best for: large organizations that already use Datadog or need a fully supported, all-in-one observability platform.
Honeycomb was built from the ground up for high-cardinality data — meaning it handles many unique attribute values (user IDs, request paths, feature flags) without forcing you to pre-aggregate or drop dimensions.2 This makes it exceptional for debugging complex microservice interactions where the signal is hidden in a rare combination of attributes.
Honeycomb's BubbleUp and heatmap visualizations let you slice traces and events in real time. It's less suited for traditional dashboarding or long-term trend analysis — it's a debugging tool first.
Best for: engineering teams that spend significant time investigating production incidents and need to ask ad-hoc questions of high-cardinality trace data.
New Relic offers a fully managed OTel experience with a generous free tier and straightforward per-GB pricing.3 It ingests OTLP directly and provides out-of-the-box dashboards, alerts, and APM without requiring you to run any infrastructure.
The platform covers traces, metrics, logs, and even browser monitoring in one product. Query performance is solid, though advanced users may find the NRQL query language less flexible than PromQL or Honeycomb's query syntax.
Best for: teams that want a managed, all-in-one solution with minimal setup and predictable per-GB pricing.
| Managed (Datadog, Honeycomb, New Relic) | Self-Hosted (Grafana LGTM) | |
|---|---|---|
| Setup time | Minutes | Days to weeks |
| Ops cost | $0 (vendor handles it) | Your team's time + infra |
| Data control | Vendor-dependent | Full control |
| Scaling | Automatic | You scale it |
| Cost at low volume | Low (free tiers exist) | Low (your own infra) |
| Cost at high volume | Can be very high | Predictable infra cost |
If your team is small or you need to move fast, go managed. If you're already running Kubernetes and have SRE bandwidth, self-hosting LGTM can save significant money at scale.
There's no single best OTel backend — the right choice depends on your scale, budget, and team. Grafana LGTM is the most flexible open-source option. Datadog is the enterprise standard. Honeycomb excels at debugging high-cardinality systems. And New Relic is the easiest all-in-one cloud solution.
Start with a managed option to reduce friction, and keep your instrumentation portable with standard OTLP so you can switch later if your needs change.
Disclosure: Some links in this article are affiliate links. We only recommend products we've researched and believe deliver genuine value. You pay nothing extra, and it helps us keep the lights on.
This page was written by the engine and the engine is still on the line. The conversation below picks up where the article stops.
Yes — the picks above are the engine's current verdicts. Ask a sharper version of this question below and you'll get a custom answer with the latest pricing.