How a Global Logistics Company Saved $6M Negotiating AWS EDP and Azure MACC Simultaneously

A multinational logistics company with operations across four continents faced a critical renewal window: both AWS and Azure contracts were expiring within the same six-month period. This synchronisation, typically seen as a logistical burden, presented a strategic opportunity that the company's IT leadership initially failed to recognise.

The company's cloud footprint was split logically and operationally. AWS hosted the core business-critical systems: route optimisation engines running on EC2 instances, fleet telematics data pipelines in Kinesis, and relational databases on RDS for real-time tracking and compliance. Azure hosted enterprise productivity tools, the Dynamics 365 ERP system for supply chain planning, and significant Azure Kubernetes Service (AKS) workloads for containerised microservices.

AWS was pushing an Enterprise Discount Program (EDP) commitment of $28M over three years. The EDP structure came with rigid conditions: no Savings Plans stacking allowance, an aggressive 80% utilisation commitment threshold, and minimal flexibility for workload migrations or decommissions. The vendor had templated the proposal with limited negotiation room.

Microsoft was simultaneously pushing a MACC (Microsoft Azure Consumption Commitment) deal bundled with an aggressive M365 E5 upgrade (including Copilot add-ons) that the business had not requested. The MACC baseline was set conservatively to protect Microsoft's revenue floor, but actual Azure consumption was running 22% below the proposed commitment level. The Copilot bundling was a margin play by the account team.

The fundamental problem: these two negotiations were treated as independent workstreams by the client's procurement team. AWS and Azure account teams operated in isolation, each believing they held the primary vendor relationship and had no visibility into the parallel renewal.

We reframed this as a single multi-cloud negotiation event rather than two independent deals. The presence of two vendors renewing simultaneously is leverage; it creates choice. Most logistics companies never exploit it.

AWS EDP Restructure

We started with a full FinOps audit across the client's 14 AWS accounts spanning production, disaster recovery, development, and data analytics environments. We identified concrete right-sizing opportunities:

• Oversized RDS instances (db.r5.2xlarge) serving non-latency-sensitive analytics workloads that could run on db.t4g instances with 40% cost reduction
• Idle EC2 Reserved Instances purchased in previous planning cycles that no longer served active workloads ($420K annual waste)
• NAT Gateway egress traffic that could be optimised through VPC endpoint restructuring (saving $840K annually)
• CloudWatch and logging overhead consuming 8% of the total bill without commensurate operational value

This FinOps work yielded $3.2M in concrete optimisation opportunities. We presented these findings to AWS not as complaints but as fact: "We're identifying optimisation opportunities worth $3.2M. The EDP commitment you've proposed must account for this optimised baseline, not today's unoptimised spend."

We then challenged the EDP structure itself. The 80% utilisation commitment is designed to lock customers into inflated baselines; if you optimise, you breach the commitment. We negotiated hard on two fronts:

Utilisation Commitment Flexibility: We secured an 18-month true-up window, meaning utilisation was measured quarterly with flexibility to adjust the baseline downward if optimisations were deployed. This removed the perverse incentive to avoid efficiency.

Savings Plans Stacking within EDP: Standard EDP terms forbid stacking Savings Plans on top of the EDP discount. We negotiated explicit permission to layer 1-year Savings Plans on compute-heavy services (the route optimisation engines). This added 7% discount on top of the EDP rate.

On AWS Graviton (ARM-based processors), we negotiated $1.2M in committed Graviton compute credits as part of the deal structure. Graviton instances are 15-20% cheaper than equivalent x86 capacity, and the route optimisation engine is architecturally suitable for ARM migration (stateless, containerised).

Azure MACC Right-Sizing

On Azure, we took a radically different approach: we challenged the baseline consumption assumptions. The client's MACC proposal was structured around a $4.1M annual consumption commitment, but actual Azure spending over the prior 12 months was $3.2M — a 22% gap.

We obtained granular Azure consumption data broken down by service and business unit. We identified:

• AKS infrastructure with significant idle capacity due to batch processing patterns that only consumed 65% of reserved node allocations
• SQL Database oversizing in non-critical reporting systems
• Unused premium features across several managed services that inflated the bill unnecessarily

The M365 E5 uplift was the real red flag. Microsoft's account team wanted to force all 3,200 users onto E5 licenses (including Copilot) at $30 per user per month. We conducted a detailed audit: which business functions genuinely required Advanced Threat Protection, Advanced Compliance, and Copilot? Only 47% of the user base (1,500 users) actually needed E5 features. The rest could remain on E3 ($22 per user per month) with selective Copilot licenses for power users.

This simple fact-based segmentation saved the client $2.4M over the contract term.

Cross-Vendor Competition Strategy

The final lever was creating explicit competitive tension. We signalled to AWS that Azure was offering better economics on the containerised workloads (which they were, due to AKS cost advantages). To AWS, we highlighted that if Dynamics 365 continued to grow, Azure's advantage would widen, potentially displacing EDP volumes.

To Azure, we noted that AWS was matching the discount terms on compute-adjacent services and that the client might consolidate ERP analytics workloads onto AWS if Azure pricing remained uncompetitive.

Neither vendor knew the other was negotiating simultaneously. When informed (tactfully, through their sales leadership), both vendors significantly improved their final offers. AWS reduced the utilisation commitment threshold from 80% to 72% and extended the true-up window. Microsoft dropped the Copilot bundling requirement entirely and offered a volume discount tier that improved economics on the smaller E5 footprint.

The negotiation concluded with $6M in verified savings across the full contract terms:

• AWS EDP Restructure: $3.6M (lower commit due to right-sizing, better discount through Savings Plans stacking, Graviton credits, improved utilisation terms)
• Azure MACC Right-Sizing: $2.4M (MACC baseline reduction, selective E5 licensing, elimination of unwanted Copilot bundling)

Under NoSaveNoPay's 25% gainshare model, the client retained $4.5M (75% of the total savings). The company deployed the FinOps optimisations and Graviton migrations in parallel with the contract transitions, ensuring the negotiated terms reflected realistic operational costs.

Both AWS and Azure account teams had projected standard renewals with minimal discounting. Neither expected coordinated leverage. The logistics company's IT leadership now understands that simultaneous vendor renewal windows are strategic assets, not just administrative burdens.