EDP NegotiationSavings Plans OptimizationReserved Instances StrategyEC2 Right-SizingS3 Cost ReductionEgress NegotiationMigration CreditsSupport Tier AdvisoryMulti-Cloud LeverageBedrock AI PricingEDP NegotiationSavings Plans OptimizationReserved Instances StrategyEC2 Right-SizingS3 Cost ReductionEgress NegotiationMigration CreditsSupport Tier AdvisoryMulti-Cloud LeverageBedrock AI Pricing
White Paper · Free Download

Serverless Cost Management Guide.

A complete buyer-side cost management playbook for AWS serverless workloads — Lambda, Fargate, DynamoDB, EventBridge, API Gateway, and Step Functions optimization, commitments, and negotiated rates.

$2.4B+
AWS spend reviewed
500+
Engagements
38%
Avg reduction
$340M+
Documented savings
Abstract

What's inside.

Serverless workloads are routinely sold as the cost-optimal architecture choice, but they create their own cost-management problems. Lambda billing is per-millisecond, DynamoDB is per-request, API Gateway is per-call, and EventBridge meters every event. At scale, the unit economics drift quickly: a single misconfigured concurrency limit, an excessive payload size, or an unbatched event pattern can move monthly costs by six figures. This guide is the cost management playbook for production AWS serverless workloads.

We cover the seven serverless services that consistently account for more than 80% of serverless spend in enterprise environments: Lambda, Fargate, DynamoDB, API Gateway, EventBridge, Step Functions, and S3 (for serverless data layers). For each service, we document the cost levers, the commitment options (where they exist), the architectural choices that compound across millions of invocations, and the negotiated-rate options available through your EDP or PPA.

The 2026 edition incorporates Compute Savings Plans for Lambda and Fargate, the latest DynamoDB Reserved Capacity strategies, the post-2024 provisioned concurrency pricing, and Step Functions Express vs Standard cost modeling at high volume. Case studies are drawn from a $4M-to-$48M annual serverless spend range across SaaS, fintech, and media customers.

Table of contents

  1. Executive summary — the cost shape of serverless at scale
  2. Lambda cost levers — memory, duration, concurrency, and Arm64
  3. Lambda commitments — Compute Savings Plans applied to Lambda
  4. Fargate cost optimization — ECS vs EKS Fargate and Spot
  5. DynamoDB pricing modes — on-demand, provisioned, and Reserved Capacity
  6. API Gateway — REST vs HTTP vs WebSocket cost comparison
  7. EventBridge — Bus vs Pipes, archive, and replay cost analysis
  8. Step Functions — Standard vs Express at high volume
  9. S3 for serverless data layers — request, storage, and transfer optimization
  10. Negotiated-rate options via EDP and PPA for serverless workloads
  11. Appendix A — Lambda right-sizing worksheet with worked examples
  12. Appendix B — serverless commitment portfolio model

Download the white paper

Submit your work email to receive the full PDF (48 pages, no marketing follow-up unless you ask). Personal email domains are not accepted.

Please use a work email. Personal email domains (gmail, yahoo, outlook, hotmail, icloud, aol, proton) are not accepted.

By submitting, you agree to receive this white paper. We do not share email addresses with AWS or any third party.

Methodology

Built on 500+ engagements.

Every figure, benchmark, and recommendation in this white paper is grounded in primary data from real AWS engagements. We have advised on more than $2.4 billion in AWS spend across 500+ engagements, spanning financial services, SaaS, media, retail, healthcare, public sector, and AI-native companies. The data set is anonymized and aggregated; no individual customer agreement is identifiable.

This white paper is buyer-side analysis. It is not affiliated with, sponsored by, or reviewed by Amazon Web Services. The recommendations reflect what works in negotiation against AWS's standard playbook, not AWS's preferred customer behavior. Where the two diverge, we have written from the customer's perspective.

Related resources

If you found this useful.

Need help applying the framework?

500+ engagements. 38% average reduction. We build your negotiation strategy within 48 hours of kickoff.