New Relic Alternatives and Competitors: What to Compare Before You Leave

Every option is described on the same seven rows: what it is, pricing shape, OpenTelemetry support, licence and lock-in, who answers an incident, what moving involves and when it wins. Pricing is described by its unit (per gigabyte, per seat, per host) from the vendors' own pages, read on 4 November 2026, not compared figure by figure. Anything we could not confirm from a vendor page is marked "not compared here" instead of estimated. AceMQ supports New Relic, Grafana, Prometheus, Elasticsearch, OpenSearch and Datadog, and resells none of them.

New Relic describes its platform as more than 50 capabilities, including APM with distributed tracing, infrastructure monitoring, digital experience monitoring (browser, mobile and synthetic monitoring, which covers real user monitoring), log management, error tracking and AIOps, billed through one model. An alternative has to replace the parts you use, not the whole list, so start from an inventory of what your teams open every week.

New Relic's bill has two main drivers. Data ingest is metered per gigabyte beyond a monthly free allowance, and full platform users are priced per seat. Core users cost less and basic users are free, so the user count that matters is the number of engineers who need the full platform, not everyone who looks at a dashboard. Add-ons such as extra retention, the EU data center and additional synthetic checks sit on top. A migration is rarely the first fix, because each of these drivers can be changed inside New Relic.

The reasons that survive that first pass are usually strategic: a wish to keep telemetry in your own account, a standardization on OpenTelemetry and open storage, a consolidation onto an engine already run for logs, or a move to host-based pricing because it forecasts better. Those are legitimate. A bill that grew because debug logs were left on is not a reason to migrate.

New Relic's pricing page (read 4 November 2026) builds the bill from two core metrics: users and data ingest, or compute and data ingest. Every account gets 100 GB of ingest a month free; beyond that, ingest is billed per gigabyte, at a higher rate on Data Plus, which adds up to 90 days of extra retention, advanced log obfuscation and enhanced export. Ingest is calculated on the data saved after processing, so drop and trimming rules reduce the bill. Full platform users are priced per seat by edition, core users cost less and basic users are free; Pro and Enterprise can instead be bought on compute with no user licences, and Enterprise is quoted by sales. New Relic does not charge per host.

The figures themselves, with a worked example, are in our blog on New Relic pricing, ingest and users explained.

Stay when agents, distributed tracing and alert conditions are in good order and the ingest bill is driven by a handful of sources. AceMQ's New Relic consulting attributes ingest by team and source, finds telemetry that nothing queries, and puts drop rules and sampling at the collection layer so the reduction holds. Typical engagements are written up in New Relic telemetry volume and cost governance and instrumentation coverage assessment. Staying also avoids rebuilding every NRQL dashboard and alert condition, which is the largest hidden cost of leaving.

The Grafana open-source stack covers the same ground in pieces: Grafana to query, visualize and alert, Prometheus or Mimir for metrics, Loki for logs, Tempo for traces and Alloy as an OpenTelemetry-based collector. There is no licence fee, and the data stays in your account. The cost moves to operations: sizing and upgrading storage, managing cardinality and retention, and keeping alert routing clean. Prometheus and the OpenTelemetry Collector are Apache 2.0, while Grafana, Loki, Tempo and Mimir are AGPL-3.0, which only matters if you modify them and offer them as a service. This is the option where AceMQ's support is most direct: see Grafana consulting and support, Prometheus consulting and support and the Grafana support options guide.

Choose it when you want OpenTelemetry-first collection and open storage and have, or will buy, the capacity to run it. It is a poor fit for a team that wants New Relic's automatic APM instrumentation and one console with no one to operate it.

If you already run Elasticsearch or OpenSearch for search or logs, their observability features avoid adding a second data platform. Elastic describes Elastic Observability as covering APM, logs and infrastructure, with OpenTelemetry-first collection and the choice of Elastic Cloud or self-managed deployment. OpenSearch is Apache 2.0 under the OpenSearch Software Foundation and documents observability features for traces and metrics. Sizing is the main risk: log and trace volume drives shard counts, retention and cost on any Lucene-based engine. AceMQ supports both: ELK stack consulting, OpenSearch consulting and Elasticsearch consulting.

Datadog is the closest like-for-like swap: another SaaS platform that runs ingestion, storage and the interface. It changes the pricing unit from users and gigabytes to hosts and usage. Infrastructure and APM are priced per host, and logs by ingest with indexing and retention priced separately; see Datadog's own pricing page for current figures. New Relic's own pricing page states that it does not charge per host, so which is cheaper depends on host count against ingest and seats. Estimate both from your own numbers before deciding. AceMQ supports Datadog estates: see Datadog consulting and Datadog support.

OpenTelemetry is the vendor-neutral way to collect traces, metrics and logs. New Relic documents first-class support and accepts OTLP data, Datadog lists OpenTelemetry support, Elastic calls its observability product OpenTelemetry-first, and Alloy is an OpenTelemetry-based collector. If agents are replaced by OpenTelemetry instrumentation and a collector you control, changing the backend becomes a routing change instead of a re-instrumentation project. Half-way estates, where some services report through vendor agents and others through OpenTelemetry, are common, and AceMQ builds the collector pipelines that unify them.

1. Attribute the bill first. Find which sources, teams and users generate the ingest and the seats. If five sources are most of it, fix those before deciding anything else.
2. List what you would have to rebuild. Count NRQL alert conditions, key dashboards, synthetic monitors and agent types. That inventory is the migration estimate.
3. Decide who operates it. A SaaS owns the platform, and a self-managed stack needs on-call engineers or a support contract. Name the owner before you choose the tool.
4. Plan the historical data. Telemetry does not migrate by itself. Keep the old account readable through your retention window, or export what compliance requires; New Relic lists enhanced export under Data Plus. Run both systems in parallel until alerts match.
5. Stay if the reasons are weak. If the only reason is a surprise invoice, an ingest audit is a smaller project than a migration.

All checked 4 November 2026. New Relic: pricing and OpenTelemetry introduction. Datadog: pricing. Grafana: open source projects, and the project licences on GitHub for Prometheus, the OpenTelemetry Collector and Grafana. Elastic: Elastic Observability. OpenSearch: documentation. The AceMQ rows are AceMQ's own description of its services.