Analyzing Your Monthly Monitoring Bill for Hidden Costs

Your observability bill tells a story — but it's written to obscure the plot. Each vendor structures billing differently, making comparison nearly impossible without deliberate decomposition.

Datadog: The SKU Maze

A typical enterprise Datadog bill has 8-15 line items:

Infrastructure ($15-23/host/month) — base layer. APM, logs, custom metrics all separate.

Log Management — three charges: ingestion ($0.10/GB), indexing (~$1.70/million events for 15-day retention), and extended retention. The gotcha: Datadog indexes by default. Without exclusion filters, you pay for indexing you may not need.

APM ($31-40/host/month) plus Indexed Spans at $1.70/million. For microservices: 50 services x $40 x 12 months = $24K/year before span costs.

Custom Metrics ($5-8 per 100 series). Cardinality makes this go from rounding error to material fast.

What to challenge: APM on batch-only hosts? Unqueried custom metrics? Unqueried indexed logs? Dev on the same plan as prod?

New Relic: Simple Pricing, Complex Reality

$0.35-0.50/GB ingested + user fees. But the ingest multiplier means 1 GB compressed = 5-10 GB metered after decompression and enrichment. Full Platform users cost $349-549/user/month — one accidental upgrade = $4,200-6,600/year.

Splunk: License Models

Three models — wrong one costs 3x more:

• Ingest-based: $150-450/GB/day. 100 GB/day = $15K-45K/month
• Workload (SVA): By compute consumed. 30-50% cheaper for efficient compression
• Entity-based: Per monitored entity. Better for infra-heavy, worse for log-heavy

Hidden costs: acceleration, summary indexing, heavy forwarder compute, premium apps.

Grafana Cloud: Three-Part Model

Component	Logs	Metrics	Traces
Processing	$0.05/GB	-	$0.05/GB
Write	$0.40/GB	$8/1K series/mo	$0.40/GB
Retention	$0.10/GB/mo	Included	$0.10/GB/mo

Cross-Vendor Comparison

Unit	Datadog	New Relic	Splunk	Grafana Cloud
Infra per host/mo	$15-23	In ingest	Varies	$0
Logs per GB	$0.10+idx	$0.35-0.50	$5-15/day/30	$0.45+ret
Metrics per 1K/mo	$50-80	In ingest	Included	$8
APM per host/mo	$31-40	In ingest	Varies	$0

Monthly Bill Review Habit

Every month, 15 minutes: total spend change? Top category shift? New line items? Cost-per-host trend? Cost-per-host climbing with flat infrastructure = telemetry volume problem.

Worked example: same workload, four vendors (illustrative)

Profile: 60 hosts, 30 GB/day logs, 25K custom series, 8K spans/min, 15 engineers.

Vendor	Modeled monthly	Biggest driver
Datadog	~$9,800	Log indexing + custom metrics
New Relic	~$7,200	Metered ingest + 8 Full Platform seats
Splunk (ingest)	~$12,500+	Daily ingest cap pricing
Grafana Cloud	~$3,400	Metrics write + log retention

What to do this week

• [ ] Build a one-page bill map: SKU → telemetry type → team owner
• [ ] Flag line items that appeared in the last 90 days (scope creep)
• [ ] Compare cost/host vs last quarter — flat hosts + rising $ = volume issue
• [ ] Cross-check with Datadog pricing guide

Sources & further reading

• New Relic pricing — ingest and user tiers
• Splunk pricing — license models
• Grafana Cloud pricing — logs/metrics/traces split

---

Related Reading

• The Observability Spend Audit
• Datadog Pricing Calculator
• Splunk Volume-Based Pricing
• New Relic Ingestion Costs
• Committed Spend Optimization for Datadog

Use the SignalCost Calculator → to model these scenarios with your own numbers.