Enterprise Discount Structures for Multi-Year Cloud Contracts (2026 Guide)

How Cloud Commitment Discounts Work

Multi-year cloud discount structures all rest on a single mechanism: you promise the provider a minimum amount of spend over a fixed term, and in return the provider applies a percentage discount to your qualifying consumption. The promise is the product. AWS, Microsoft Azure and Google Cloud each package it differently — the Enterprise Discount Program, the Microsoft Azure Consumption Commitment, and Committed Use Discounts respectively — but the economics are the same, and so are the ways a buyer can get the structure wrong.

The reason this matters is scale. These are not line-item discounts on a single SKU; they apply across most of your consumption for the entire term. A two-percentage-point swing in the discount on a $5M annual commitment is $100,000 a year, or $300,000 across a three-year deal. The difference between a well-structured commitment and a poorly structured one is rarely the headline discount — it is the term, the ramp, the eligible-spend definition, and the flexibility clauses that determine whether you actually capture the discount you signed up for.

This guide is the structural reference for our enterprise cloud contract negotiation cluster. It covers all three hyperscalers because most large enterprises now run at least two of them, and the right commitment strategy is a portfolio decision — not three separate negotiations run in isolation.

One distinction matters before going further. A commitment discount is not the same as a reserved-capacity discount. Reserved Instances and Savings Plans on AWS, reservations on Azure, and resource-based CUDs on Google Cloud all discount a specific compute footprint — a particular instance type or a fixed hourly rate. A commitment discount, by contrast, sits above the whole account and discounts qualifying consumption regardless of which services produce it. The two stack: an enterprise can hold Savings Plans covering its steady-state compute and an EDP discounting everything on top. Treating the account-level commitment as a substitute for capacity-level reservations — or vice versa — leaves money on the table on both layers.

AWS: The Enterprise Discount Program

The AWS Enterprise Discount Program (EDP) is the most mature of the three structures and the template most buyers know. Under an EDP, you commit to a minimum total spend on AWS over a one-to-three-year term and receive a percentage discount applied to qualifying consumption. AWS does not publish its discount tiers, so benchmark data is the only reliable reference a buyer has.

The practical eligibility trigger is current or projected annual AWS spend of roughly $1M–$3M. Above $3M, an EDP negotiation is standard practice; above $10M, AWS routes the account to its large-enterprise and strategic teams, which unlocks deeper discounts and more flexible commercial structures. Observed discounts run from around 5% at entry level to 15–20% at the largest commitments. As a working benchmark, a $1M annual commitment typically returns 6–9%, and the curve steps up at recognisable spend milestones — around $1.5M, $2M and $5M — rather than scaling smoothly.

Term is the other axis. At the same spend level, a three-year EDP is usually worth around 15% where a one-year deal returns roughly 10% — a five-percentage-point term premium for accepting a longer lock-in. We cover the negotiation mechanics in detail in our AWS enterprise agreement negotiation guide and benchmark the discount-by-spend curve in our AWS EDP discount benchmarks breakdown. For the distinction between a standard EDP and a custom Private Pricing Agreement, see AWS EDP vs Private Pricing Agreement.

Azure: The Microsoft Azure Consumption Commitment

The Microsoft Azure Consumption Commitment (MACC) is Azure's equivalent: a contractual commitment to spend a defined amount on Azure over a period, usually tied to an Enterprise Agreement or a Microsoft Customer Agreement. Microsoft offers an Azure commitment discount in recognition of the committed spend level, and that discount tier is contingent on the size of the commitment. Unlike AWS, Microsoft rarely frames the MACC discount as a standalone number — it is negotiated inside the broader Microsoft commercial relationship, which means Azure commitment, Microsoft 365 and licensing leverage interact.

The MACC's defining advantage is the breadth of what counts toward it. Many purchases through the Microsoft Commercial Marketplace — including enrolled third-party ISV software — draw down the committed spend, and the 2026 MACC expansion widened that eligibility further by making more ISV products qualify. For a buyer, this changes the maths: software you were going to buy anyway can be routed through the marketplace to retire your Azure commitment, making an apparently aggressive commitment number far more achievable. Where the Azure relationship is primarily about the Microsoft commercial agreement rather than pure consumption, our AWS vs Azure enterprise agreement comparison sets out how the two structures differ in practice.

Google Cloud: Committed Use Discounts

Google Cloud's Committed Use Discounts (CUDs) come in two forms. Resource-based CUDs commit to a specific quantity of compute (vCPUs and memory) in a region. Spend-based flexible CUDs — the more relevant structure for a multi-year commitment strategy — commit to a minimum hourly dollar amount and apply a discount across machine families, regions and services such as GKE and Cloud Run.

Google publishes its CUD percentages, which makes it the most transparent of the three. On the compute flexible commitment that spans Compute Engine, GKE Autopilot and Cloud Run, spend-based flexible CUDs deliver 28% for a one-year term and up to 46% for a three-year term — a far steeper term premium than AWS or Azure typically offer. As of January 2026, Google fully migrated spend-based CUDs to a "multiprice" direct-discount model, where the discount applies directly to SKU prices rather than appearing as a separate billing credit. That makes the saving more transparent in billing, but it does not change the underlying commitment risk: you are still promising a minimum hourly spend for the full term.

The Three Programmes Compared

The structures rhyme, but the details that determine your downside differ sharply. The table below summarises the 2026 picture for a buyer weighing a multi-year commitment across providers.

Term Economics: One Year vs Three

The term decision is where most of the value — and most of the risk — concentrates. Longer terms carry materially higher discounts because the provider values the certainty of locked revenue. On AWS, the step from a one-year to a three-year EDP is worth roughly five percentage points; on Google Cloud's flexible CUDs the gap between the 28% one-year rate and the 46% three-year rate is far larger. The vendor is paying you to give up flexibility, and the size of that payment tells you how much they value the lock-in.

The buyer's question is not "which discount is bigger" — it is "how confident am I in my spend three years out?" A three-year commitment is only valuable if you can consume against it for three years. The term premium is real money, but it is conditional money: it only materialises if you avoid the shortfall. An enterprise with a stable, growing cloud footprint and a clear migration roadmap should capture the three-year premium. An enterprise facing a possible divestiture, a workload-repatriation programme, or an uncertain AI-infrastructure trajectory should treat the longer term with caution, or structure it so the committed floor stays conservative.

Work the term premium as a number, not a feeling. If a one-year EDP returns 10% on a $5M spend and a three-year deal returns 15%, the extra five points is worth $250,000 a year — provided you consume the full commitment in each of the three years. Now weigh that against the downside: if Year 3 spend falls 30% short because a business unit is divested, the shortfall true-up can erase a full year of the premium and more. The three-year term is the right call when the probability-weighted upside of the premium clearly exceeds the probability-weighted cost of a shortfall. Quantifying both sides — rather than reaching for the bigger headline number — is what separates a disciplined commitment from an expensive one.

The five-year structure

Some enterprises are now offered five-year structures, typically as two sequential three-year terms with pricing committed upfront. These carry the deepest headline discounts, but they compound the forecasting problem: the further out the commitment runs, the wider the band of plausible spend outcomes, and the greater the chance the committed floor diverges from reality. A five-year commitment is appropriate only where the technology roadmap is genuinely stable and the flexibility clauses below are strong enough to absorb a change of plan.

The Overcommitment Trap

The single most expensive mistake in multi-year cloud commitments is overcommitting. Vendor account teams are incentivised to maximise the committed number, and opening commitment proposals routinely sit 15–30% above the customer's realistic spend curve. That gap looks harmless when the deal is signed and the discount is celebrated — and then it converts into a shortfall true-up at term end.

The arithmetic is unforgiving. Commit to $10M a year and consume $8M, and you lose the discount on the $2M shortfall — effectively paying the full published rate on spend you never made. Depending on your discount percentage, that penalty runs $200,000–$500,000 in additional cost. The discount you negotiated so hard to win is handed straight back, plus interest, because the commitment was modelled on the vendor's optimism rather than your actual trajectory.

This is not a hypothetical. In one engagement, a mid-market software company was about to sign a $4.2M EDP with an aggressive Year 1 ramp. Independent modelling identified a 28% shortfall risk against the proposed commitment profile. Restructuring the ramp before signing — lower in Year 1, stepping up as deployment caught up — removed roughly $840,000 of penalty exposure while preserving the discount. The lesson is consistent across providers: the commitment floor must be modelled from your own spend curve, not accepted from the vendor's proposal.

Flexibility Clauses That Protect You

The discount is the headline; the flexibility clauses are where an experienced negotiator actually earns their fee. None of the following are standard terms — each is a negotiated provision, and each materially reduces the risk that a multi-year commitment turns into a penalty.

Ramped commitment

Structure the commitment to rise as you actually deploy — lower in Year 1, increasing in Years 2 and 3 — rather than committing the full annual figure from day one. A ramp matches the commitment to the real adoption curve and is the most effective single defence against shortfall. If the vendor pushes for a larger total commitment, the ramp is the concession to demand in return.

Rollover and usage banking

By default, cloud commitments are use-it-or-lose-it against each period. Negotiate the right to carry unused commitment forward — usage banking — so an underspent quarter is not simply lost. Some enterprise deals allow at least one-period carryover or partial rollover at renewal; press for it explicitly, because nothing in the standard contract grants it.

Reallocation rights

If one project consumes less than forecast, you should be able to repurpose that budget elsewhere. Negotiate the right to shift unused commitment across services, regions or business units, so a single underperforming workload does not create a shortfall while another part of the business overspends at full rate.

Marketplace drawdown

Confirm in writing which spend counts toward your commitment. Eligible AWS Marketplace spend can add an estimated 5–15% of headroom against an EDP commitment, and Azure's MACC umbrella is wider still — many third-party ISV purchases through the Microsoft Commercial Marketplace draw down the committed spend. Routing software you were buying anyway through the marketplace can be the difference between meeting and missing your floor.

Cure periods and shortfall remedies

Negotiate a cure period before any shortfall true-up is charged, giving you time to accelerate eligible spend rather than paying a penalty on day one of the shortfall. Pair it with a defined remedy — for example, the right to extend the term to absorb the unused commitment instead of forfeiting the discount.

A Portfolio View: Committing Across Two Clouds

Most enterprises that run multi-cloud treat each provider commitment as an isolated event, negotiated by whichever team owns that platform. That is a mistake. The commitments interact, and managing them as a portfolio changes the maths. A workload that is genuinely portable between AWS and Azure is a competitive lever on both — the credible ability to shift spend is what moves an account team from a standard discount to an escalated one. An enterprise that signs a deep, long commitment to one provider quietly surrenders that leverage everywhere else.

The portfolio question is allocation: how much of your total cloud spend to lock to each provider, and for how long. Over-concentrating commitment in one cloud maximises that provider's discount but removes your ability to grow the other without paying full rate. A balanced approach keeps a defensible committed floor with the primary provider — capturing the bulk of the available discount — while leaving enough uncommitted headroom to grow a second provider opportunistically and preserve competitive tension at every renewal. The marketplace-drawdown rules reinforce this: because Azure's MACC umbrella is wider than AWS's, software spend you control can be steered to retire whichever commitment is harder to hit, smoothing the whole portfolio.

This is also where timing compounds. Staggering commitment renewals so they do not all fall in the same quarter means you always have at least one live competitive alternative when negotiating any single deal. Enterprises that let every commitment expire together lose that — and walk into each renewal with no credible threat to move spend. For the cross-provider negotiation framework, see our enterprise cloud contract negotiation pillar and the AWS vs Azure enterprise agreement comparison.

Setting Your Commitment Floor

The commitment floor is the number you can defend regardless of what happens to your roadmap — the spend you are confident you will make even in a conservative scenario. Everything above that floor should be earned with additional discount, additional flexibility, or both. The discipline is to start from your own bottom-up forecast, apply a margin of safety, and treat the vendor's proposed number as an opening position to be modelled against, never as the baseline.

To pressure-test a proposed commitment against modelled discount ranges and a defensible floor for each provider, use our cloud commitment discount calculator. For the deeper AWS-specific playbook — including the negotiation sequence, benchmark tables and flexibility-clause language — download the AWS EDP Negotiation Playbook, and review our existing AWS EDP negotiation guide for the end-to-end process. If you are weighing a commitment this year, you can request a confidential briefing and we will model your floor against current benchmark data before you respond to the vendor.