AWS Contract Negotiation: European Broadcast Media Group Saves €5.1M

The Challenge

A European broadcast and streaming media group with approximately 4,500 employees operated free-to-air and streaming services across six European markets. The organisation held a significant and growing AWS estate, with €8.3M in annual committed spend across EC2, S3, and specialised services including AWS Elemental for broadcast media encoding. Yet the group had never consolidated its AWS commercial relationship into a unified strategy.

Instead, individual teams purchased Reserved Instances independently based on near-term capacity requirements. This decentralised approach created multiple misalignments: Reserved Instance types no longer matched actual instance utilisation patterns (teams had changed their workload architecture but not their RI purchases); the portfolio contained no Compute Savings Plans to provide flexibility across instance families; and the organisation had not negotiated an Enterprise Discount Program despite annual spend exceeding €8M.

Enterprise Support was costing €220K per year at standard percentage-based rates. The fragmented buying pattern also meant that S3 storage tiers were not optimised—significant volumes of cold archive data remained in standard S3 when Glacier Deep Archive would have delivered 60 to 70 percent cost reduction.

Individual teams purchased Reserved Instances independently, creating a fragmented portfolio where many RIs no longer matched actual workload architecture. There was no EDP, no Compute Savings Plans, and no consolidated commercial strategy.

The Approach

Redress began with a comprehensive audit of the current AWS commercial state across all six markets. The assessment mapped every Reserved Instance purchase against actual running workloads, identified which RIs were applied to instance types no longer in use, and quantified the total committed spend eligible for an Enterprise Discount Program negotiation.

The engagement then proceeded through four parallel workstreams. First, we rationalised the Reserved Instance portfolio by consolidating all RIs into a smaller set of flexible, frequently-used instance types, then released the mismatched instances. Second, we modelled Compute Savings Plans across a three-year commitment term as a replacement for 74 percent of the legacy RI portfolio, offering teams greater flexibility to change instance families and sizes as workloads evolved.

Third, we prepared the commercial case for an Enterprise Discount Program negotiation. We documented the €8.3M annual committed spend, outlined the organisation's growth trajectory across both free-to-air and streaming pipelines, and signalled AWS that migration to a consolidated EDP framework would simplify the purchasing process and allow teams to adopt new AWS services more rapidly without individual purchasing delays.

Fourth, we benchmarked Enterprise Support pricing against actual service consumption. The analysis revealed that the organisation's support case load and TAM engagement patterns did not justify the high cost of premium support tiers. We negotiated a fixed-fee annual support arrangement that provided the same service layer as before but at a stable, lower cost.

In parallel, we conducted a storage audit identifying approximately 2.5 PB of cold archive data sitting in S3 Standard when Glacier Deep Archive would have been appropriate. We modelled the migration strategy and estimated the annual savings.

The Outcome

The consolidated AWS commercial strategy delivered €5.1M in negotiated savings over the initial three-year EDP commitment term. The savings broke down as follows:

• Enterprise Discount Program (€2.3M): A negotiated 16 percent discount on the €8.3M annual committed spend, locked in for three years, provided the largest lever. This discount applied to EC2, RDS, and a growing slate of newer services the organisation planned to adopt.
• Reserved Instance Rationalisation and Compute Savings Plans Conversion (€1.6M): By retiring mismatched RIs and replacing 74 percent of the remaining portfolio with Compute Savings Plans, we reduced the annual RI spend by approximately €533K while increasing flexibility. Teams could now change instance families and sizes within the Savings Plan commitment without incurring additional costs.
• Enterprise Support Negotiation (€700K): A fixed-fee annual support agreement of €110K (down from €220K per year at percentage-based rates) locked in support costs across the commitment term, delivering €110K annual savings.
• S3 Storage Class Optimisation (€500K): Migration of the 2.5 PB archive cohort to Glacier Deep Archive, combined with lifecycle policy adjustments, generated approximately €167K in annual savings.

The changes also delivered non-financial benefits. The consolidation of Reserved Instances into a smaller, more strategic set reduced administrative overhead in RI management. Compute Savings Plans reduced the risk of RI waste as workloads evolved. The EDP framework simplified future procurement of new services—teams no longer needed individual purchasing justifications for small new workload pilots, as they fell within the committed EDP spend envelope.

The fixed-fee support agreement stabilised the support budget and removed the perverse incentive (common in percentage-based support models) for vendors to inflate customer case volumes to drive higher support revenue.

Key Takeaways

Consolidation Unlocks Leverage: The single largest savings opportunity arose from consolidating a fragmented buying pattern into one unified commercial negotiation. Individual teams purchasing RIs independently forgo the scale discount that comes from a consolidated commitment. At €8M+/year of committed spend, enterprise-grade EDP discounts in the 15–18 percent range are typical.

RIs Become Liabilities When Workloads Shift: A common trap in RI management is purchasing RIs with long commitment terms, then discovering that the workload has changed and the RI no longer applies. This client had purchased RIs three years ago for instance types that were no longer used. Compute Savings Plans, with more flexibility across instance families, reduce this friction.

Support Tiers Should Track Usage, Not Spending: Enterprise Support pricing is often presented as a percentage of total spend. But the actual value of support is determined by case volume, severity distribution, and TAM engagement. For many large customers, moving to a fixed-fee support arrangement below the percentage-based rate is possible if the case load and TAM profile do not justify premium tier pricing.

Storage Class Optimisation Is Predictable But Often Overlooked: Archive and cold storage tiers routinely sit one or two storage classes higher than necessary, especially when purchased as part of application rollout without deliberate retention and migration policies. A simple audit can often identify 10–20 percent of storage spending that can be optimised downward with minimal operational friction.

Commercial Strategy Should Guide Architecture Decisions: When teams purchase RIs independently, they optimise for their local budget cycles, not for enterprise efficiency. A unified commercial strategy that makes EDP discounts and Savings Plans the preferred procurement path encourages teams to choose architectures that cost less and remain more flexible.