For enterprises managing multi-cloud environments — or evaluating which cloud commitment to prioritise — understanding the commercial architecture of each hyperscaler's enterprise programme is essential. AWS EDP, Azure MACC (Microsoft Azure Consumption Commitment), and Google Cloud Commit are all designed to exchange long-term spend commitment for discounts, but they differ significantly in structure, flexibility, and what's included.
This guide is part of our AWS enterprise negotiation series and complements our dedicated guides on Azure committed spend negotiation and GCP cost optimisation. For the broader context on using cloud programmes as leverage against each other, see our cloud vendor negotiation leverage guide and cloud enterprise discount negotiation guide.
The three enterprise cloud programmes have meaningfully different structures. Azure MACC is credit-based and the most Marketplace-friendly. AWS EDP offers the best stacking with service-level commitments (RIs/SPs). GCP Commit is the most negotiable on individual service discounts. Each also varies significantly in how it handles AI/GenAI services — increasingly the fastest-growing spend category.
Programme Overview: Three Different Models
AWS EDP (Enterprise Discount Program)
AWS EDP is a committed-spend programme where you commit to spend a minimum dollar amount on eligible AWS services over 1–5 years. In return, AWS applies a percentage discount to eligible spend on your bill. The discount is applied as a bill credit rather than a pre-reduction of list prices. EDP stacks on top of service-level commitments (Savings Plans, Reserved Instances). Full details: AWS EDP negotiation playbook.
Azure MACC (Microsoft Azure Consumption Commitment)
Azure MACC (previously Azure Monetary Commitment) works differently from EDP. Rather than a discount applied to your bill, MACC is a credits purchase — you pay Microsoft upfront (or agree to pay) for Azure credits, which are then consumed against your Azure usage at list prices. The "discount" comes from the negotiated credit value: e.g., paying $4.5M upfront to receive $5M in Azure credits represents an effective 10% upfront discount. MACC also has Marketplace implications: ISV software purchased through Azure Marketplace can count toward burning down your MACC balance — a significant operational advantage. Full details: Azure MACC negotiation guide.
GCP Commit (Committed Use Discounts + Enterprise Agreement)
Google Cloud's enterprise commitment structure has two components: (1) service-level Committed Use Discounts (CUDs) that apply to specific compute resources, and (2) a private pricing agreement (often called a GCP Commit or Google Cloud Agreement) that provides a discount on total or category-level spend for larger commitments. GCP's enterprise programme is generally considered the most negotiable on a per-service basis, and Google has been increasingly aggressive in offering EDP-equivalent arrangements to win workloads from AWS. Full details: GCP cost optimisation guide.
Side-by-Side Contract Term Comparison
| Feature | AWS EDP | Azure MACC | GCP Commit |
|---|---|---|---|
| Programme type | Bill discount on eligible spend | Pre-purchased credits at discounted rate | Hybrid: per-service CUDs + programme discount |
| Minimum entry (2026) | ~$1M/year | ~$1M/year | ~$1M/year |
| Typical term | 1–5 years (3yr most common) | 1–5 years (3yr most common) | 1–3 years |
| Discount mechanism | % discount on eligible spend | % discount on credit purchase price | % discount on eligible spend |
| Stacking with service commitments | Yes — excellent (SP + RI + EDP) | Yes — via AHB and RIs (different structure) | Yes — CUDs + programme discount |
| Marketplace applicability | Negotiable (ISV Accelerate) | Strong — by default for many ISVs | Limited |
| AI/GenAI service coverage | Negotiable (Bedrock, SageMaker) | Yes (Azure OpenAI, Copilot) | Yes (Vertex AI, Gemini) |
| Ramp structure | Negotiable | Standard ramp available | Negotiable |
| Shortfall treatment | Invoice at shortfall (negotiable) | Credits expire (different risk) | Invoice at shortfall (negotiable) |
| Credit rollover | N/A (discount model) | Usually annual expiry (negotiable) | N/A (discount model) |
| Price escalation cap | Negotiable | Credits locked at purchase price | Negotiable |
| Migration incentives | MAP credits (separate) | Azure migration programme | Google migration programme |
| Support integration | Separate (percentage-based) | Included via Unified Support | Separate (negotiable) |
Discount Rate Benchmarks by Programme
Achievable discount rates depend heavily on commitment size, term, competitive dynamics, and negotiation quality. The following benchmarks represent well-negotiated outcomes for each programme:
| Annual Commitment | AWS EDP (achievable) | Azure MACC (achievable) | GCP Commit (achievable) |
|---|---|---|---|
| $1M | 8–12% | 7–11% | 8–14% |
| $3M | 12–16% | 11–15% | 12–18% |
| $5M | 14–18% | 13–17% | 15–20% |
| $10M | 17–21% | 15–19% | 18–24% |
| $20M | 20–24% | 18–22% | 22–28% |
| $50M+ | 24–28% | 22–26% | 26–32% |
GCP typically offers higher headline discount rates because Google is more aggressive in winning new workloads. However, the effective discount depends on what spend is included in the programme — GCP's Marketplace applicability is weaker than Azure's, which may reduce the real-world benefit for organisations with significant third-party software spend.
AI and GenAI Service Coverage: A Critical 2026 Difference
AI and GenAI services are now among the fastest-growing cost categories for enterprise cloud users. How each programme handles these services differs significantly — and this difference can be worth millions annually.
| AI/GenAI Service | AWS EDP Coverage | Azure MACC Coverage | GCP Commit Coverage |
|---|---|---|---|
| Foundation models (LLM APIs) | Bedrock: Negotiable | Azure OpenAI: Yes (usually) | Vertex AI / Gemini: Yes |
| ML training and inference | SageMaker: Negotiable (SP available) | Azure ML: Yes | Vertex AI Training: Yes |
| AI-powered productivity tools | Q Business: Negotiable | Microsoft 365 Copilot: via EA (separate) | Google Workspace AI: separate |
| Vector database / RAG services | OpenSearch, Aurora pg_vector: Yes | Azure AI Search: Yes | AlloyDB, Spanner: Yes |
| Custom model fine-tuning | Bedrock fine-tuning: Negotiable | Azure OpenAI fine-tuning: Yes | Vertex AI fine-tuning: Yes |
The key takeaway: Azure MACC has the most consistent AI service inclusion by default. AWS EDP requires explicit negotiation to include Bedrock and AI services. GCP includes AI services but its commitment structure may not optimally match token-based consumption patterns. For AI-first organisations, this is a significant factor in programme selection.
Marketplace Dynamics: A Major Structural Difference
The Marketplace treatment is perhaps the most significant structural difference between the three programmes for enterprises with significant third-party software spend.
Azure Marketplace and MACC
Azure MACC has the strongest Marketplace integration of any hyperscaler programme. ISV software purchased through Azure Marketplace by default counts toward burning down your MACC credits. This means your enterprise software portfolio — ServiceNow, Snowflake, Databricks, and dozens of others available on Azure Marketplace — can be funded through MACC credits. This is a powerful procurement centralisation mechanism. Combined with Microsoft's large ISV partner network, Azure Marketplace + MACC is often the most attractive option for enterprises with diverse software estates.
AWS Marketplace and EDP
AWS Marketplace has excellent reach but weaker default EDP integration. Marketplace spend does not automatically count toward EDP commitments or receive EDP discounts. This requires explicit negotiation, as covered in our AWS Marketplace private offers guide. When negotiated, AWS Marketplace EDP applicability can be excellent — but it requires deliberate effort and is not as seamless as Azure's approach.
GCP Marketplace and Commit
GCP Marketplace has the weakest default integration with enterprise commitment programmes. Third-party software purchased through GCP Marketplace does not typically count toward GCP Commit commitments. This is a material limitation for enterprises with significant ISV software spend who want to consolidate cloud and software procurement.
Using Each Programme as Leverage Against the Others
The single most effective tactic for improving enterprise cloud contract terms is credible competitive tension. Each programme operates most generously when the buyer is genuinely evaluating alternatives. Our cloud vendor negotiation leverage guide covers the full framework; here are the key multi-cloud leverage dynamics:
AWS vs Azure leverage
Microsoft has strategic reasons to prevent Azure workloads from migrating to AWS — and vice versa. If you have significant Microsoft 365, Teams, and Azure workloads, you have built-in leverage: AWS cannot easily offer Microsoft productivity stack alternatives, but they can offer infrastructure cost advantages. Use detailed cost-of-migration analysis to demonstrate the economic viability of workload movement as a genuine negotiating tool, not a bluff.
AWS vs GCP leverage
GCP has consistently been willing to offer above-market discounts to win significant workloads from AWS. If your workloads are technically compatible with GCP (particularly data analytics, ML, and containerised applications), a credible GCP proposal is one of the most effective levers for improving AWS EDP terms. Google's aggressive posture on winning AI/ML workloads in 2025–2026 makes this particularly effective for AI-forward organisations.
Azure vs GCP leverage
GCP can create competitive pressure for Azure data analytics and AI/ML workloads (BigQuery vs Azure Synapse / Azure AI). Microsoft responds aggressively to GCP competition on data platform and AI workloads. This is a viable lever for improving Azure MACC terms for data-heavy organisations.
The most powerful position is genuine multi-cloud capability — not just the appearance of it. Enterprises that have actually built competency in two hyperscalers achieve materially better contract terms than those who use competitive proposals purely as a bluff. The hyperscalers know the difference.
Contract Flexibility Comparison
| Flexibility Dimension | AWS EDP | Azure MACC | GCP Commit |
|---|---|---|---|
| Ramp structure (lower Year 1) | Negotiable (common) | Standard ramp available | Negotiable |
| Mid-term renegotiation | Uncommon but possible | More common (credits can be topped up) | Possible with Google account team |
| Spend category reallocation | High (any eligible service) | Very high (most Azure services) | Moderate (CUDs are service-specific) |
| Exit / termination rights | Very limited by default (negotiable) | Limited (credits paid upfront) | Very limited (negotiate T4C) |
| Change of control provisions | Negotiable (important) | Part of Microsoft EA framework | Negotiable |
| Shortfall rollover | Negotiable | N/A (credits expire) | Negotiable |
Which Programme Is Best for Your Situation?
Profile: 80%+ workloads on AWS, some Microsoft 365 and Azure for Microsoft stack. $8M/year AWS spend.
Recommendation: Focus on AWS EDP as primary programme. Negotiate AI service inclusion (Bedrock), Marketplace applicability for key ISVs, and Support cost cap. Use Azure MACC (even a modest $500K) as negotiating leverage to improve EDP terms. Ensure Savings Plans and RIs are optimised before EDP to accurately model committed spend.
Profile: Heavy Microsoft 365, Dynamics 365, and Azure workloads. $5M Azure, $1M AWS. Multiple ISV software platforms on Azure Marketplace.
Recommendation: Azure MACC as primary programme — maximise Marketplace credit burning and AI service coverage. Negotiate MACC credit rollover and ramp structure. Use AWS EDP (even small) and GCP evaluation as leverage for MACC terms. Coordinate MACC negotiation with Microsoft EA renewal for maximum leverage.
Profile: $3M cloud spend growing 80%/year. Heavy ML/AI workloads. Primary need is AI service flexibility and growth-friendly commitment structure.
Recommendation: Consider GCP Commit for AI/ML workloads — Vertex AI coverage is strong. Negotiate a short-term (1-year) commitment with renewal options rather than locking into a 3-year structure during rapid growth. Use GCP terms as leverage for AWS Bedrock pricing and EDP AI inclusion. Keep 30–40% of spend uncommitted to maintain flexibility.
Profile: Financial services or healthcare. $20M total cloud spend split $12M AWS / $6M Azure / $2M GCP. Strong data sovereignty and compliance requirements.
Recommendation: Negotiate all three programmes, but treat each as leverage for the others. AWS EDP should target 20–24% at this spend level. Azure MACC should include government/regulatory cloud regions in scope. GCP commitment should be small but genuine — it provides the most competitive pressure against AWS and Azure without requiring significant commitment. Engage independent advisory support given the complexity.
10 Cross-Cloud Contract Negotiation Tactics
Initiate formal enterprise programme discussions with AWS, Azure, and GCP simultaneously, within a 30-day window. All three will become aware of the competitive process. This is the most powerful signal you can send to a cloud account team — that you are actively evaluating alternatives.
Create a standardised comparison framework: effective discount on total spend (including exclusions), AI service coverage, Marketplace applicability, flexibility provisions, and support terms. Present this framework to each cloud vendor and ask them to complete it. This discipline creates comparable data and forces each vendor to address their weaknesses directly.
GCP consistently offers migration incentives — architecture reviews, engineering support, and credits — to organisations considering migrating workloads from AWS or Azure. Use these offers as concrete leverage. "GCP has offered us $500K in migration support and a 22% programme discount — what can you match?" is a powerful negotiating statement.
Each cloud vendor's fiscal calendar creates timing leverage: AWS (calendar year), Microsoft (June fiscal year), Google (calendar year). The most powerful position is having all three negotiations active simultaneously around your preferred signing window. See our software renewal timing strategy guide for the detailed framework.
AI/GenAI spend is projected to represent 20–40% of enterprise cloud budgets by 2027. Negotiate explicit inclusion of AI services (Bedrock, Azure OpenAI, Vertex AI) in your programme before signing any multi-year commitment. Without this, your fastest-growing cost category may sit entirely outside your programme discount framework.
All three hyperscalers have demonstrated willingness to raise list prices during the term of enterprise commitments. Negotiate price protection provisions: caps on annual list price increases for covered services, or commitments to maintain current pricing for your committed service portfolio. This is particularly important for 3–5 year terms.
If possible, negotiate all three cloud enterprise commitments to expire in the same quarter. This creates maximum leverage at renewal — you can negotiate all three simultaneously, using each as leverage for the others, rather than having staggered renewals that reduce your cross-cloud negotiating position.
M&A activity affects cloud commitments significantly. AWS, Azure MACC, and GCP Commit all have provisions that allow the vendor to renegotiate or terminate on a change of control. If M&A activity is possible in your planning horizon, negotiate explicit change-of-control provisions in all three agreements — covering both scenarios: you being acquired and you acquiring others.
The information asymmetry between a buyer who negotiates cloud commitments once every 3 years and a hyperscaler account team that negotiates daily is extreme. An independent advisor with current benchmark data across all three programmes typically improves outcomes by 3–8 percentage points per programme — and helps ensure the non-discount terms (flexibility, AI coverage, Marketplace, exit rights) are appropriately negotiated.
When evaluating competing proposals, build a total value model that includes: programme discount rate × eligible spend, credits and migration incentives, Marketplace applicability value (ISV spend that can count toward commitment), support cost treatment, flexibility provisions value, and AI service coverage value. A 1% lower programme discount is easily worth accepting if it comes with better Marketplace applicability and AI service coverage. See our cloud cost optimisation firm rankings for advisors that can build this model.
Navigate Multi-Cloud Contract Negotiations
Our top-ranked cloud negotiation advisors help enterprises run AWS, Azure, and GCP negotiations simultaneously — with current benchmark data and outcome-based options.