Datadog vs New Relic
New Relic is the closest like-for-like Datadog alternative: a managed, all-in-one observability suite. The difference is the meter — New Relic charges for data ingested plus users, not per host. Here's when that flips the math in your favour.
The short version
Both are turnkey, fully-managed platforms covering APM, infrastructure, logs, and more. If you're leaving Datadog because per-host, per-module billing is unpredictable but you still want a hosted suite (not an open stack to run yourself), New Relic is the natural switch. Its usage-based model — 100GB/month free, then per-GB ingest — rewards teams with many hosts but modest data. The catch is the per-user fee: teams with lots of "full platform" users can pay more than expected.
How they compare
| Datadog | New Relic | |
|---|---|---|
| Model | Turnkey suite | Turnkey suite |
| Pricing basis | Per-host + per-module | Per-GB ingested + per-user |
| Free tier | 5 hosts, 1-day | 100GB/mo + 1 full user |
| Paid from | ~$15/host (infra) + APM etc. | ~$0.30–0.40/GB; full user ~$99/mo |
| Self-host | No | No |
| OpenTelemetry | Ingests (custom-metric billing) | Supported |
| Breadth | Broadest catalog, RUM/synthetics/security | Full platform, 50+ capabilities |
| Best when | Stable usage, want max breadth | Many hosts, modest data, few power users |
When New Relic wins
- You have many hosts but modest data volume — paying per-GB instead of per-host is far cheaper.
- You want a managed suite, not an open stack to assemble and operate.
- Most of your team are read-only / core users — those are free or cheap, only full users cost.
When to stay on Datadog
- You need the broadest integration catalog and the deepest RUM, synthetics, security and DB monitoring.
- You have many full-platform users — New Relic's per-user fee can erase the per-GB savings.
- Your data volume is very high — at some point per-GB ingest overtakes per-host, so model both against real numbers.
Migration effort
Moderate. New Relic uses its own agents, so you swap the Datadog agent for the New Relic agent (or point OpenTelemetry at New Relic's OTLP endpoint). Historical data doesn't transfer — rebuild dashboards and alert policies, run both in parallel for a release cycle, then cut over. Tie deploys into whichever you keep so regressions map to the exact release — Buddy ships a native Datadog action, and New Relic exposes a deployment-markers API for the same purpose.