Dynatrace vs New Relic
Dynatrace, an Austrian company, maps an estate automatically from one agent per host and bills on consumption per host hour and per GiB. New Relic, an American one, bills per GB ingested plus per full-platform user, takes OTLP natively and offers a free allowance. Large, poorly documented estates lean towards Dynatrace; container fleets with a few engineers lean towards New Relic.
Vendors can pay for visibility on this site; it never changes what we say about a product. How that works.
Choose Dynatrace ifnobody can fully describe your estate any more, you want the dependency graph and root-cause suggestions generated for you, and a formal procurement and implementation project is acceptable.
Choose New Relic ifyou run many small containers, want OpenTelemetry instrumentation that stays yours, have few people needing full access, and prefer to start on a free allowance instead of a sales call.
Dynatrace and New Relic side by side
| Dynatrace | New Relic | |
|---|---|---|
| Our application performance monitoring rank | #10 of 15 | #11 of 15 |
| Established in | Austria | United States |
| Pricing | Consumption-based per host hour or memory-GiB hour and per GiB ingested, published rates | Free tier of 100 GB a month; then per GB ingested plus per core and full-platform user, published |
| Free tier | — | Yes |
| Open source | — | — |
| Self-hostable | — | — |
| Built for | Large estates nobody fully documented, where auto-discovery pays | Container fleets where host counting would be punishing |
| Not for | Small teams that want a fixed monthly price they can forecast | Large groups of occasional viewers who all need full access |
Both are reviewed in full in our application performance monitoring ranking: Dynatrace and New Relic.
Where they differ
Dynatrace 1 · New Relic 2 · open 1- 01
Hosts versus gigabytes
Dynatrace counts host hours alongside GiB ingested, and per-host metering suits stable virtual machines while punishing container fleets. New Relic dropped host pricing for data plus users, which suits many small containers but punishes a large group of occasional viewers needing full seats. The better meter is decided by your estate and headcount, not by the vendor.
Depends on the buyer
- 02
Automatic discovery
Dynatrace's OneAgent draws the dependency graph without manual work, surfacing services nobody remembered, and its root-cause suggestions beat the market average. Our data makes no such claim for New Relic, whose listed strengths are pricing model and OTLP. The caveat is that an auto-generated map still needs someone to name services and owners.
Edge: Dynatrace
- 03
Owning your instrumentation
New Relic treats OTLP as a first-class input, so code instrumented with OpenTelemetry can move to another vendor later. Dynatrace ingests OTLP as well, yet the product is built around OneAgent and its good features assume it. If you expect to change tools one day, New Relic leaves you with less to redo.
Edge: New Relic
- 04
Getting started without procurement
New Relic has a genuinely usable free allowance and published pricing, so a team can try it on real services first. Dynatrace publishes its rates too, but it has a learning curve and is sold and implemented in a way that expects a procurement process. Small teams are not its buyer; they are closer to New Relic's.
Edge: New Relic
When neither fits
Neither fits an estate that must keep every trace unsampled or host the monitoring itself. IBM Instana traces every request with no sampling by default and has a self-hosted edition, and Splunk AppDynamics serves large Java and .NET estates with a self-hosted controller.
Dynatrace or New Relic: questions
3 answersIs Dynatrace or New Relic better for containers?
New Relic's data-plus-users pricing suits a fleet of small containers, where counting hosts would be punishing. Dynatrace's consumption model includes host hours, and per-host metering is a poor match for container fleets. Dynatrace's auto-discovery is still valuable if the estate is large and undocumented.
Which is easier to forecast, Dynatrace or New Relic pricing?
Dynatrace's consumption pricing across hosts, ingest and retention is difficult to forecast before you run it. New Relic has two meters, but full-platform seats cost many times a basic seat. With New Relic, model the seat mix first; with Dynatrace, measure real telemetry volume first.
Do Dynatrace and New Relic both accept OpenTelemetry data?
Yes, both ingest OTLP. New Relic accepts it natively as a first-class path next to its own agents. Dynatrace's strongest features assume OneAgent. Ask each vendor during the trial which features stop working when you send data through your own collector.