New Relic

What New Relic actually is

New Relic is a full-stack observability platform: application performance monitoring, infrastructure monitoring, browser and mobile monitoring, synthetics, log management, error tracking, AIOps, and dashboards over one telemetry data store. It is one of the oldest companies in this market — founded in 2008 by Lew Cirne (now executive chairman; Ashan Willy is CEO), headquartered in San Francisco, IPO'd in December 2014, and taken private in November 2023 when Francisco Partners and TPG bought it for roughly $6.5 billion. Wikipedia records a revenue milestone of around $1 billion announced in 2025. Whatever else you think of the pricing, this is a mature vendor with a decade of public-company discipline behind it.

The product thesis is consolidation: instead of stitching APM, logs, infrastructure, and front-end monitoring from separate tools, you put all telemetry into one platform and query it with NRQL. The catalog runs to 50+ capabilities — distributed tracing, errors inbox, workloads, service level management, cloud cost intelligence, change tracking, and, more recently, AI-assisted investigation under the "Intelligent Observability" banner. In our view the unified query model is the real reason to buy; the capability count is marketing arithmetic.

The other defining trait is pricing architecture. New Relic charges for two things — data ingest and user types — and explicitly not per host, per agent, per container, or per function. That makes cost track telemetry volume rather than infrastructure size, which favors chatty microservices fleets and penalizes teams that log everything. Whether that trade works for you is the central question of this review.

Pricing: what you really pay

Four editions, all sharing the same 100 GB/month free ingest allowance:

• Free — $0, no credit card. 100 GB of data ingest per month, one free full platform user, unlimited basic users, automatic logs obfuscation, access to all 50+ capabilities. The catch is firm: exceed 100 GB in a month and data ingest stops and you lose platform access until you upgrade or the next month begins. An 85% warning email and in-platform alert precede this.
• Standard. Full platform users at $10 for the first user and $99 for each additional (maximum 5); core users at $49/user; basic users $0. Ticketed support with a 2-business-day response SLA and SAML SSO.
• Pro. Full platform users at $349/user on annual commitment or $418.80/user month-to-month; core users $49/user. Unlimited full platform users, commitment and savings-plan options, Data Plus eligibility, and a 2-hour critical initial support response SLA.
• Enterprise — contact sales. FedRAMP Moderate and HIPAA eligibility (with Data Plus), priority ticket routing, 1-hour critical response SLA.

Data beyond the free 100 GB costs $0.40/GB on the original option or $0.60/GB on Data Plus (longer retention, more compute-backed querying, and compliance eligibility), with EU data center storage at an extra $0.05/GB per month. Optional Advanced Compute capabilities (Transaction 360, Security RX, New Relic AI) are metered per Compute Capacity Unit on Pro and Enterprise and can be toggled off to cap spend; there is also a full consumption model with no user licenses at all.

Our reading of this model. The free tier is the most generous in the category — one full user with the entire platform, forever, no card — and the Standard entry ($10 first user) is a genuinely cheap way in. The pain arrives at Pro: $349 per full user per month is serious money, and teams where "everyone needs a full seat" feel it immediately. The mitigation is honest tiering — basic users at $0 and core users at $49 cover dashboards, alerts, and code-level visibility without full seats — but it requires discipline about who gets which seat. Data at $0.40–$0.60/GB is competitive with per-host APM pricing as long as you control ingest; the fastest way to blow this budget is unfiltered logs.

Performance & infrastructure

New Relic is SaaS with regional data residency: accounts store data in the US or EU (EU at the noted $0.05/GB premium), and there is no self-hosted edition — the Enterprise tier adds compliance scope, not your own servers. Agents cover the usual runtimes (Java, .NET, Node, Python, PHP, Go, Ruby), plus infrastructure agents, browser and mobile SDKs, synthetic monitors, and open standards ingestion (OpenTelemetry-compatible data is accepted alongside native agents).

We have not benchmarked New Relic, and there is nothing to run our sysbench/fio/iperf3 suite against here (see /benchmarks for the VPS numbers we do publish). What matters for performance is the agent overhead question every APM buyer asks: sampling rates and ingest filters are configurable, and the pricing model gives you a direct lever — every metric and log line you drop is money saved at $0.40/GB. Treat ingest configuration as cost engineering, not just performance tuning.

On reliability, the public record is the strongest argument: a platform running since 2008 at roughly $1 billion revenue scale, used across enterprises that put it in their incident path. The contract record is narrower — availability commitments exist for Pro and Enterprise editions with full users and are defined in your Terms; on Standard (or without full users) those commitments are explicitly vacated and New Relic only promises commercially reasonable efforts. That is unusually candid language, and it means Standard-tier buyers should not assume an uptime SLA.

One capability worth singling out is logs in context: because logs, traces, and metrics share one data store, an APM transaction can pull the exact log lines written during it without a separate logging vendor or correlation IDs stitched across systems. For debugging production incidents this is the feature teams miss most when they leave, and it is the clearest practical payoff of the single-data-platform design. Just remember it is also the feature that can quietly drive ingest past your free 100 GB the week you turn it on everywhere.

Where it falls short

Cost complexity is the recurring complaint, and it is earned. Between ingest tiers, three user types, per-edition feature gates, Advanced Compute metering, and optional consumption pricing, forecasting a bill takes real work — and the jump from Standard to Pro full-user pricing ($99 to $349) is steep enough to force seat audits. Vendor-lock in cuts the other way: NRQL, custom dashboards, and alert definitions are portable only with effort.

The platform's breadth is also its learning curve. Teams that need "APM plus a few alerts" will find themselves navigating a product built for observability programs, and the AI-era additions (New Relic AI, Service Architecture Intelligence) arrive with new metering attached. We would disable Advanced Compute features until their value is proven on your workloads.

Ownership by private equity since 2023 is a neutral-to-watch item rather than a red flag — revenue and product investment appear healthy by public milestones — but buyers should keep an eye on packaging changes, which is where PE-owned infrastructure vendors tend to move first. The Standard-tier SLA language already shows how terms can differ by edition.

Who should pick this

Our pick for engineering organizations that want one platform for APM, infrastructure, logs, and front-end monitoring and are willing to police ingest to keep bills sane. The free tier is our default recommendation for individual developers and small teams who want full-platform visibility at zero cost — one full user and 100 GB is a real working setup. Standard at $10 for the first full user is the cheapest credible path to SSO and ticketed support we know of in this market.

We would look elsewhere for simple uptime monitoring (it is overkill), for teams allergic to seat management, or for organizations that require self-hosted telemetry storage. Open-source stacks (Prometheus/Grafana/Loki) cost engineering time instead of dollars and keep data in-house; per-host APM vendors cost more at small scale and less at large, chatty scale. Run the numbers with your actual ingest volume before committing to Pro seats.

Key takeaways

• Founded 2008 by Lew Cirne, San Francisco; IPO 2014, taken private by Francisco Partners and TPG for about $6.5 billion in November 2023.
• Free tier is the strongest in the category: 100 GB ingest/month, one full platform user, unlimited basic users, no credit card — but exceeding 100 GB cuts off access until the next month or an upgrade.
• Data beyond free costs $0.40/GB (original) or $0.60/GB (Data Plus), plus $0.05/GB for EU storage; no per-host or per-container charges.
• Full platform users cost $10 first / $99 each on Standard (max 5) and $349/user on Pro annual — seat tiering is where the budget goes.
• Uptime commitments apply to Pro/Enterprise full users and live in your Terms; Standard buyers get "commercially reasonable efforts" only.
• Support response SLAs are published per edition: 2 business days (Standard), 2-hour critical (Pro), 1-hour critical (Enterprise).

Sources

• https://newrelic.com/pricing
• https://newrelic.com/info/pricing
• https://newrelic.com/about
• https://docs.newrelic.com/docs/licenses/license-information/referenced-policies/service-level-availability-commitment
• https://en.wikipedia.org/wiki/New_Relic
• https://wikimili.com/en/New_Relic