A global media company spending $22M annually on AWS had renewed its Enterprise Discount Programme (EDP) on auto-pilot for three successive years. A specialist cloud cost advisory engagement identified $6M in annual savings through EDP renegotiation, Reserved Instance portfolio reconstruction, and architectural rightsizing — without any service disruption.
This is what structured advisory looks like.
Start 9 months out. Every week of lead time is leverage recovered.
A global media and entertainment company had migrated the majority of its infrastructure to AWS between 2019 and 2022, driven by streaming platform growth. By 2025, AWS annual spend had reached $22M, comprising a mix of EC2 compute, S3 storage, CloudFront CDN, and a growing RDS footprint. The company had an active AWS Enterprise Discount Programme (EDP) with a three-year committed spend agreement that was approaching renewal.
The cloud infrastructure team had grown organically, with individual teams provisioning resources independently. A centralised FinOps function was notionally in place but lacked tooling, governance authority, and bandwidth to address the scale of optimisation opportunity. AWS had recommended its own Cost Optimisation Assessment, which the advisory team found to be significantly conservative — understating the available savings to protect AWS revenue.
The CFO engaged a specialist AWS cost optimisation advisor after a peer at another media company described achieving 30%+ AWS cost reductions through a structured programme.
Editorial note: Identifying details anonymised. Savings represent the annualised run-rate reduction achieved by month six of the engagement, verified by the client's finance team against AWS billing data. Advisors referenced are drawn from our ranked AWS advisory firms.
The advisory assessment identified $6.8M in annual optimisation opportunity across six categories:
| Category | Finding | Annual Saving |
|---|---|---|
| EC2 Reserved Instances | RI coverage at 34%; Graviton instances not utilised | $2.1M |
| EDP Renegotiation | Below-market discount tier; Azure leverage available | $1.8M |
| Idle and Oversized Resources | 18% of EC2 instances <5% utilisation; 23% oversized | $1.1M |
| S3 Storage Tiering | 68% of objects never accessed after 30 days; no lifecycle policies | $620K |
| Data Transfer / Egress | Cross-region replication patterns creating avoidable egress | $480K |
| RDS Rightsizing | Dev/test RDS instances running 24/7; Multi-AZ in non-production | $380K |
The advisory team structured the programme into four workstreams running concurrently, with the EDP renegotiation on a separate commercial track:
The existing EDP was reviewed against AWS's published commitment tiers. The client's $22M annual spend qualified for a materially higher discount tier than their current agreement reflected — a gap created by three years of organic growth without commercial renegotiation. The advisory team supplemented this with a documented Azure migration assessment covering the media workload, which AWS's enterprise account team treated as a credible competitive threat. The renegotiated EDP secured an additional $1.8M annual discount improvement on the new three-year commitment.
The RI audit revealed that 34% of EC2 spend was covered by Reserved Instances, vs an optimal target of 70–75% for stable workloads. Many existing RIs were for older instance families (M4, C4) while the production fleet had moved to M6 and C6. A portfolio reconstruction plan was developed, exchanging legacy RIs via AWS's exchange programme, adding Compute Savings Plans for workload-flexible coverage, and migrating 40% of qualifying workloads to Graviton3 instances — achieving average 20% compute savings on migrated instances.
An automated inventory scan identified 847 EC2 instances with CPU utilisation below 5% over a 30-day observation period. Of these, 312 were confirmed as genuinely idle (decommissioned workloads retained "just in case") and terminated. An additional 218 instances were downsized by one instance size tier following CPU/memory profiling. Rightsizing was implemented through a tag-based governance workflow that required team leads to approve any instance above the recommended size.
S3 lifecycle policies were implemented across 14 active buckets, transitioning objects to S3 Infrequent Access at 30 days and Glacier Instant Retrieval at 90 days for non-CDN content. Cross-region replication patterns were reviewed and rationalised — three redundant replication rules that duplicated content already served via CloudFront were eliminated. Dev/test RDS instances were placed on automatic shutdown schedules, reducing runtime from 8,760 hours/year to approximately 2,500 hours.
The EDP renegotiation was the single highest-value activity in the programme and illustrates the commercial dynamic that many AWS customers miss. AWS's EDP discount tiers are based on committed annual spend, but AWS does not proactively offer customers improved tiers when organic growth moves them into a higher qualifying bracket. Customers must actively renegotiate.
The advisory team's approach to AWS EDP negotiation followed a structured framework. First, the client's trailing 12-month spend was documented and benchmarked against AWS's published and unpublished tier thresholds. Second, competitive alternatives were developed — not hypothetically but as a documented evaluation. The Azure migration assessment covered four core workloads and included a credible timeline and migration cost estimate. Third, the assessment was shared with AWS's enterprise account team as part of a formal commercial review, framing the EDP renewal as a competitive decision rather than a routine renewal.
AWS's response was a revised EDP proposal that closed approximately 70% of the identified discount gap. The advisory team negotiated the remaining 30% via a combination of incremental committed spend (a $3M annual increase to $25M) and marketplace credits that could be applied against specific AWS service spend.
Beyond the immediate savings, the engagement established a FinOps operating model to prevent re-accumulation of waste. The framework included three core components.
First, a cost allocation taxonomy with mandatory tagging on all resources above $500/month. Untagged resources were automatically flagged for review. This created accountability at the team level and enabled accurate chargeback to business units for the first time.
Second, a monthly RI and Savings Plan coverage review cadence, maintained by the FinOps function, to ensure commitment coverage kept pace with workload growth. An automated alert triggered when EC2 on-demand spend in any account exceeded $50K/month without corresponding commitment coverage.
Third, an architectural review gate for any new deployment exceeding $10K/month — requiring a documented cost estimate and rightsizing justification before provisioning. This gate, lightweight by design, captured the majority of cost-inefficient architectural decisions before they were provisioned rather than after.
We knew we had waste — every engineering team does. What we didn't know was the scale, or that our EDP was significantly under-tiered for our actual spend. The programme paid for itself in the first two months.
— VP of Engineering, Global Media Company (anonymised)This case illustrates several principles that apply broadly to enterprises with significant AWS spend.
For cloud spend above $2M annually, specialist cloud cost advisory consistently delivers 10:1 or better ROI. See our rankings of the best AWS negotiation firms and our comprehensive AWS cost optimisation guide. For the broader cloud cost optimisation framework, our cloud cost optimisation pillar guide covers all three hyperscalers. Download the cloud contract negotiation white paper for the full commercial playbook.
Overpaying on AWS? Get a free cloud spend assessment.