New Relic bills on two independent axes: data ingested, measured per GB, and user seats, billed per month. Neither axis is capped, and neither moves in isolation; a busier release cycle drives up ingestion, a new hire drives up seats, and both show up on the same invoice.
At low volume this is barely noticeable. Past a certain point, usually somewhere around 5 -10 engineers and a few hundred GB/month, the two variables start compounding, and the bill stops tracking with intuition.
This post breaks down what each New Relic tier actually includes and where the two pricing axes compound hardest. It also shows what a single-axis, data-only pricing model looks like by comparison.
Key takeaways
- New Relic bills on two independent axes: data ingested (per GB) and user seats (per month). Neither is capped, and both scale with growth.
- A 15-person team (5 Full Platform & 10 Core users) at 500 GB/month lands around $2,395/month, or ~$28,740/year combined data and user costs.
- The free tier is capped at 100 GB/month; full-stack instrumentation (APM, logs & infra) on even one moderate-traffic app can exceed that within days.
- Standard tier maxes out at 5 Full Platform users; crossing that threshold forces a move to Pro at $349 – $418.80/user/month the steepest cost jump in the pricing structure.
- Middleware prices on data ingestion only, with no per-seat charge, which removes one of the two variable cost drivers entirely.
NewRelic unethical billing
byu/jaxn indevops
The math worth running
Model a mid-sized team on New Relic’s Pro tier: 5 Full Platform users and 10 Core users, ingesting 500 GB/month.
| Features | Mid-Size Team | Enterprise Scale |
|---|---|---|
| Full Platform users | 5 × $349 = $1,745 | 25 × $349 = $8,725 |
| Core users | 10 × $49 = $490 | 15 × $49 = $735 |
| User cost subtotal | $2,235/month | $9,460/month |
| Data ingested (raw) | 500 GB | 4,900 GB |
| Free allowance | 100 GB | 100 GB |
| Billable data | 400 GB | 4,800 GB |
| Rate per GB | $0.40/GB | $0.60/GB (Data Plus) |
| Data cost subtotal | $160/month | $2,880/month |
| Total (list price) | $2,395/month (~$28,740/year) | ~$12,340/month (~$148,000/year) |
“Our monthly spend with NewRelic surpassed the value we were getting out of it. The specific incident that triggered us to look at other vendors was when we turned on JVM-level telemetry and were not aware that it incurred a “custom event” cost, which skyrocketed our bill 10X for a month.”
John D’Emic, Co-Founder and CTO at Revenium
Mid-size total excludes retention upgrades, EU data-center surcharges, or overage on synthetic checks. At enterprise scale, most organizations negotiate a custom contract and typically land 20 – 40% below list on annual commitments above $100K but the underlying two-axis structure (seats × rate, plus GB × rate) doesn’t change, only the multipliers do.
What New Relic actually is
New Relic is a full-stack observability platform APM, infrastructure monitoring, log management, distributed tracing, browser/mobile monitoring, and synthetic checks, all under one data pipeline. Every signal (metrics, logs, traces, events) flows into a single pool, metered per gigabyte, with a separate per-seat charge layered on top for full platform access. Querying runs through New Relic’s own query language, NRQL. The platform itself is mature and full-featured; the complexity lives entirely in how the two billing axes interact as usage scales.
What you actually get at each tier
| Tier | Cost | Data Ingest | Full Platform Users | Best For |
| Free | $0/month | 100 GB/month included | 1 free user | Solo developers, early experimentation |
| Standard | From $10/mo (1st user) + $99/additional user | 100 GB free, then $0.30–$0.40/GB | Max 5 users | Small teams with predictable, modest data volume |
| Pro | $349/user/year (or $418.80/mo billed monthly) | Same per-GB rate as Standard | Unlimited users | Growing teams needing more users and priority support |
| Enterprise | Custom pricing | Custom, includes Data Plus options | Unlimited, custom rates | Regulated industries, large-scale, compliance-heavy deployments |
Every tier also throws in unlimited free Basic users (view-only dashboards) and lets you add Core users at $49/month if they need query access but not full configuration rights. Once you’re past the free 100 GB, data is billed at $0.30–$0.40/GB on the standard option, or up to $0.60/GB if you need Data Plus, which buys you longer retention and compliance features like HIPAA and FedRAMP eligibility.
Where the pricing model gets complicated
“Pay for data and users” sounds simple until you look at how each cost driver actually behaves in practice:
- The free tier disappears quickly – 100 GB sounds generous until APM, log forwarding, and infrastructure monitoring are running on even one moderate-traffic app. Many teams cross the limit within the first week.
- The Standard-to-Pro jump is steep – Standard caps out at 5 Full Platform users. One more user past that forces a move to Pro at $349 –$418.80/user/month a jump from roughly $406/month (5 users on Standard) to over $2,000/month for 6 users on Pro.
- Logs drive most of the overage – Verbose logging alone can generate 50+ GB/month, and the default agent configuration forwards everything unless it’s manually filtered.
- Two variables, not one – Data costs and user costs are additive rather than bundled, so total spend can move more than expected in any given month particularly in microservices environments where telemetry volume is inherently uneven.
- Retention isn’t included by default – Standard retention is 8 days for most data types; 30/60/90-day retention requires the pricier Data Plus option.
“Middleware’s pricing model incentivizes us to send more data, which means we get better results. This was very important for us because other tools would penalize us for higher data usage, leading to worse outcomes.”
Rangaraj Tirumala, Founding Engineer at Hotplate
Why teams evaluate alternatives
None of this points to a weak product New Relic remains one of the more complete observability platforms available. The issue is that cost scales with team size and system complexity rather than with the value extracted from the data. As headcount grows, microservices multiply, or logging gets more verbose for better debugging, spend can outpace budget forecasts. A few signals it’s worth comparing options:
- Engineering time is going into tuning telemetry volume for cost control rather than using the data itself.
- Adding one engineer to the on-call rotation produces a measurable jump in the next invoice.
- Forecasting next quarter’s observability spend is closer to guesswork than budgeting.
Also read about New Relic alternatives
New Relic vs. Middleware
| New Relic | Middleware | |
| Pricing axes | Two data ingested (GB) + user seats | One data ingested (GB) only |
| Free tier | 100 GB/month, 1 Full Platform user, permanent | 14-day trial, unlimited data ingestion |
| Full Platform user cost | $10–$418.80/user/month depending on tier | $0 every user gets full access |
| Core user cost | $49/user/month | N/A no user-based tier |
| Data ingest rate (list) | $0.40/GB (Original), $0.60/GB (Data Plus) | ~$0.30/GB |
| Cost predictability | Low two independent variables compound | High cost tracks one metric (data volume) |
| Retention | 8 days default; 30–90 days requires Data Plus upgrade | Included, no separate paid tier for extended retention |
| Best fit | Teams needing NRQL, 700+ integrations, deep enterprise support tiers | Teams where headcount is growing and per-seat costs are the main pain point |
Running the same scenario on both
Take the 15-person team from earlier 5 Full Platform users, 10 Core users, 500 GB/month and price it out on each model.
New Relic (Pro tier): $2,235/month in seats + $160/month in data overage = $2,395/month
Middleware: No seat charge for any of the 15 engineers. Data cost is 500 GB × ~$0.30/GB = ~$150/month
That’s roughly a 93% reduction in this specific scenario driven almost entirely by removing the per-seat charge, since seats made up $2,235 of New Relic’s $2,395 total. The gap narrows for very small teams (1–2 users, where New Relic’s per-seat cost is minimal) and widens for larger ones, since every additional engineer adds a fixed cost on New Relic but nothing extra on Middleware.
Where Middleware differs
Middleware’s pricing model is built around a single premise: cost should scale with data volume, not headcount. There’s no per-seat charge, every engineer gets full access, and the bill tracks actual telemetry ingested rather than the number of people using the platform. That removes one of New Relic’s two variable cost drivers entirely, which makes monthly spend considerably easier to forecast as teams scale.
Revenium, an API monitoring and monetization platform, migrated from New Relic to Middleware after being blindsided by an undisclosed “custom event” charge that spiked their bill 10X in a single month. Beyond fixing the billing surprise, the switch cut their observability costs by over 10X compared to their highest New Relic bill, improved operational efficiency by 75%, and sped up incident root cause analysis by 50%.
“We primarily chose Middleware because New Relic became prohibitively expensive, considering that we were only using a subset of its features. Middleware offers feature parity with New Relic for the features that matter to us”
John D’Emic, Co-Founder and CTO at Revenium
Middleware Slashes Revenium’s Observability Costs by Over 10X
What that looks like in practice
- No per-user fees – Every engineer gets full access to Basic/Core/Full Platform tiering to manage or budget around.
- A single pricing lever – Cost is driven by data ingested at a flat rate, rather than two independent variables compounding against each other.
- Meaningful savings past a handful of seats – Teams that would otherwise hit New Relic’s per-seat cost cliff tend to see the largest gains here.
- A 14-day free trial with unlimited data ingestion – Real usage patterns can be measured before committing to a plan.
- One platform, one bill – Metrics, logs, and traces under a single pricing structure.
For teams where headcount and log volume are both growing, comparing a data-only pricing model against New Relic’s current invoice is usually a quick way to see where the numbers diverge.
Conclusion
New Relic’s pricing isn’t broken, it’s built on two variables instead of one. That’s why the bill stops feeling predictable past roughly 5–10 engineers or a few hundred GB/month, seats and data don’t just add up, they compound.
That doesn’t make New Relic weak, it’s still a full-featured platform. But cost scaling with headcount instead of data value is a real planning problem for growing teams.
Removing one axis doesn’t just cut the bill, it makes it forecastable. If seats are already your biggest New Relic line item, that’s your signal.
FAQs
What's the difference between Core, Basic, and Full Platform users on New Relic?
Basic users get free, view-only access to dashboards. Core users ($49/month) can query data and troubleshoot but can’t configure alerts or platform settings. Full Platform users have complete access, including configuration, and are the most expensive tier.
Why does my New Relic bill fluctuate so much month to month?
Because your total cost is the sum of two independent, variable inputs data ingested (GB) and user seats, small changes in either (a noisy release, a new hire, extra log verbosity) can meaningfully move your total bill.
Does Middleware support the same core observability features as New Relic?
Middleware covers metrics, logs, and traces with dashboards, alerting, and full-stack visibility the core observability workflows most teams rely on day to day without the per-user licensing model.
How do I estimate what I'd actually save by switching?
Start by pulling your current monthly data ingestion volume and user count from New Relic. Since Middleware’s pricing is usage-based only, removing per-seat costs from the equation is usually where the biggest, most immediate savings show up especially for teams with more than 5 – 6 full platform users.
