Methodology
We reviewed anonymized AWS billing data from 180 startups ($8k–$420k monthly spend) over a 12-month window. Each account was tagged against 24 common waste signals (idle resources, right-sizing gaps, savings-plan coverage, NAT Gateway misuse, etc.). Remediation outcomes were measured 90 days post-intervention to separate real savings from rebounds.
Key findings
The signals that shaped the rest of the analysis — each one backed by the full methodology in the download.
34%
Average bill reduction after 90 days
Median savings once the top three waste patterns were addressed per account.
$1,840
Monthly NAT Gateway waste per account
Routing chatty internal services through NAT Gateways was the single most common silent drain.
17%
Of spend was on idle or zombie resources
Non-production environments left running nights and weekends accounted for most of it.
$0.023
Per GB saved by switching to Intelligent-Tiering S3
For accounts with >50 TB of cold-leaning data, the move paid for itself within a single billing cycle.
71%
Had zero Savings Plans coverage
Leaving 20–35% instant savings on the table for workloads already running steadily.
Inside the report
A scannable chapter guide. Each chapter includes data tables, charts, and practitioner commentary in the full PDF.
Each pattern scored by prevalence, typical monthly impact, and difficulty to remediate.
Graviton migration, burstable instance misuse, and container density as three high-impact levers.
S3 class selection, EBS gp3 migration, and snapshot hygiene benchmarked against industry defaults.
NAT Gateway, inter-AZ traffic, and VPC endpoint strategies with actual before/after invoices.
A staged commitment model so you don't over-lock-in while still capturing discounts.
Why untagged resources cost startups roughly 9% more and how to retrofit without breaking IaC.
A stepwise triage sequence for containing runaway spend within 48 hours.
Full report
All the data, charts, and appendices — no gate, no drip campaign, no retargeting pixel circus.