You are registered. Your playbook is ready. Read the full 2026 edition of the AWS EDP Negotiation Playbook below.
Prepared by Atonement Licensing · buyer-side advisory · last reviewed June 2026. Figures are list-level or clearly labelled indicative ranges. The $12M-per-year AWS estate used below is a representative benchmark scenario for illustration, not a quote.
Executive summary
Size the AWS Enterprise Discount Program commitment to the spend you control, and every other term in the deal falls into place behind it. The discount AWS offers is real, but it is priced against a committed floor you pay whether or not you consume it, and the first proposal always stretches that floor as far as your data can be made to reach. The gap between the vendor's growth curve and your defensible forecast is where the money is won or lost, not in the headline percentage.
On a representative enterprise spending about $12M a year on AWS, an opening three-year EDP proposal built on the account team's growth curve models a committed floor near $45M across the term, front-loaded so the early years already sit above today's run rate. The same business, sized bottom up from high-confidence workloads and back-loaded, models a defensible floor near $33M — leaving roughly $12M of committed spend resting on growth that may not arrive. Because the discount applies to consumption above the floor as well, that $12M buys no extra discount; it buys only shortfall risk. The decision in front of procurement is how low to set the floor, how to shape the ramp, what counts toward it, and which protections to place before signature.
This playbook covers the full negotiation: how AWS constructs an EDP offer and where the risk actually sits, the twelve levers in the order that protects you, the six-month commit-sizing timeline, when a Private Pricing Agreement beats a flat discount, how Marketplace retirement and data transfer pricing work, the shortfall and overcommit mechanics, and how to hold the renewal to your own forecast. Every section ends with the action a buyer should take and names the contract mechanism that makes it stick.
How AWS builds an Enterprise Discount Program offer, and where the risk sits
An EDP is a private agreement that layers a cross-service percentage discount on top of your AWS consumption in exchange for a committed minimum spend over a fixed term, most often three years. The discount is published nowhere. It is set by the size of your commit, the length of the term, your growth trajectory, and how well you negotiate. Two companies with identical annual spend can sign materially different deals on structure alone.
The commitment is where the risk concentrates. You agree to a minimum spend, usually rising year over year through a ramp. If actual usage falls short of the committed floor in a period, you owe the difference as a true-up under the shortfall language in the agreement. That single clause is more expensive than any discount is valuable, which is why sizing discipline beats percentage chasing in every deal we have run.
How the account team scores your deal
Inside AWS, an EDP proposal is evaluated on total committed dollars, year-over-year growth of the commitment, and term length. The discount your account team can offer expands as those three inputs rise, which is why every first proposal pushes all three at once. Knowing the scoring logic tells you what each concession costs the seller: a longer term is cheap for them to reward, a flat commitment is expensive, and shortfall protections cost them almost nothing when your consumption record is strong.
Pricing beyond standard authority routes through a private pricing desk, not the account manager, and turnaround on revised terms is measured in weeks, not days. AWS runs a fiscal year ending December 31, and quarter ends sharpen the field's flexibility. Build that latency into your timeline so the final quarter of the negotiation is not compressed into the final week, where pressure favours the side whose deadline matters less. Your renewal date matters to you; the AWS quarter end matters to them.
The three documents that govern the deal
The EDP sits on top of the AWS Customer Agreement and the AWS Service Terms, with the EDP addendum carrying the commitment, the discount schedule, and the shortfall mechanics. Read all three together. The addendum controls the money, but eligibility questions, service exclusions, and Marketplace treatment trace back to the Service Terms. Ask for the current versions in writing before you negotiate, and keep the executed set with your contract record.
Action. Before you respond to any EDP proposal, decompose it into floor, ramp, term, and discount, and treat the discount as the reward for certainty about your own consumption, never as a reason to commit beyond it.
The slice of the benchmark $45M proposal that rests on unproven growth, removed by sizing the floor to high-confidence consumption and back-loading the ramp (indicative).
Consumption above the committed floor earns the same negotiated discount with none of the shortfall exposure, which is why the rational floor sits at the bottom of your confidence range, not the middle.
The EDP levers, sequenced: commit, ramp, term, Marketplace, and discount last
Discount is one lever of twelve, and it is the weakest place to spend negotiating power. The levers below are ordered by how much they protect the buyer over a full term. Settle each structural term before you let the conversation move to the percentage, because once the percentage is agreed the account team has no reason left to concede structure.
| Lever | What it does | When it works best |
|---|---|---|
| 1. Commit sizing | Sets the floor at confident consumption | Always; the floor is the real risk |
| 2. Ramp shape | Back-loads growth into later years | When adoption builds over the term |
| 3. Term length | Trades a longer term for a deeper discount | When the cloud roadmap is stable |
| 4. Marketplace retirement | Counts eligible Marketplace spend against the commit | When you buy third-party software on AWS |
| 5. Private pricing (PPA) | Sets service-specific rates below standard | When one or two services dominate spend |
| 6. Egress and data transfer | Negotiates data transfer rates directly | When egress is a large, growing line |
| 7. Savings Plans interplay | Stacks rate discounts inside committed spend | When compute is steady state |
| 8. Shortfall protection | Caps or softens the true-up exposure | When the ramp could outrun the forecast |
| 9. Credits and migration funds | Adds onboarding or migration funding | When a workload move is planned |
| 10. Renewal and co-term | Aligns renewal timing across agreements | When agreements renew apart |
| 11. Exit and flexibility | Builds room for divestiture or decline | When the business may contract |
| 12. Discount | The headline percentage, last | After every structural term is set |
The sequence is the strategy. A buyer who opens with the percentage signals that structure is conceded, and the account team will hold the floor and the ramp exactly where the first proposal put them. A buyer who opens with commit size, ramp shape, and shortfall language forces the structural conversation while the seller still has a quota reason to stay at the table.
Migration funding through the AWS Migration Acceleration Program is negotiated alongside the EDP, not inside it. If a data centre exit or a VMware estate move is on your roadmap, table it during the EDP negotiation and ask for MAP credits in the same commercial package. Account teams can combine the two when the buyer asks and rarely volunteer it when the buyer does not.
Action. Write your asks down in this order and refuse to discuss the discount until the structural levers above it are settled in the draft.
The discount is the reward for certainty about your own consumption. It is never a reason to commit beyond it.
Sizing an AWS commit in the next two quarters? Our advisors run this playbook with you, buyer side only.
Cloud Contract NegotiationThe six-month commit-sizing timeline and where the saving comes from
The work that decides whether an EDP saves money happens before you accept a number. Sizing is a forecasting exercise, not a negotiating tactic, and it must be built bottom up from workload plans rather than top down from last year's bill plus a growth percentage. The timeline below is the sequence we run with clients.
Forecast
Build a bottom-up spend forecast by workload and account, separate committed from discretionary spend, and sum the high-confidence tier into the committable floor.
Structure
Model the ramp against the real adoption curve, inventory Marketplace-eligible purchases, and draft the shortfall, carry-forward, and exit terms before any price talk.
Close
Benchmark the discount against commit size and term, anchor with a written term sheet, and time the signature to an AWS quarter end or the December 31 year end.
| Months before signing | What to do | Why it matters |
|---|---|---|
| 6 to 5 | Build a bottom-up spend forecast by workload and account | The commit must rest on demand you can defend |
| 5 to 4 | Separate committed spend from discretionary and experimental spend | Only commit to the spend you control |
| 4 to 3 | Model the ramp against the real adoption curve | A floor that rises faster than usage becomes a penalty |
| 3 to 2 | Benchmark the discount against commit size and term | Know the trade before AWS frames it for you |
| 2 to 1 | Negotiate structure first: floor, ramp, shortfall, Marketplace | Lock protection before price |
| 1 to 0 | Close near an AWS quarter end or the December 31 year end | Timing pressure favours the prepared buyer |
Building a forecast that holds up
A defensible forecast is assembled workload by workload, not as one blended number. For each material workload, record current monthly spend, the owning team, the growth or retirement plan, and a confidence rating. Sum the high-confidence tier and that is your committed base. The medium tier informs the ramp. The speculative tier, the proofs of concept and the unfunded initiatives, stays out of the commitment entirely, because the EDP discount still applies to consumption above the floor.
The chart below shows the shape that protects a buyer. The figures are an illustrative index, with the final-year floor set to 100, not a market benchmark.
A back-loaded ramp keeps the early floors under the adoption curve, where the shortfall risk is highest. You do not need to commit to spend for the discount to apply to it; you need the commitment only to earn the rate. Spend above the floor gets the same discount with none of the shortfall risk, so the rational floor sits at the bottom of your confidence range. Let the account team argue you upward against your own evidence, and make every increment of floor buy a visible structural concession in return.
Action. Start six months out and spend the first three months on the forecast, not the negotiation. The cheapest concession AWS ever grants is the one you never need because the floor was right.
4Private pricing and rate cards: when service rates beat a flat discount
For organizations with concentrated spend, a Private Pricing Agreement can be worth more than the headline EDP percentage. A PPA sets negotiated rates on specific services, typically the ones that dominate the bill: EC2 compute, S3 storage, CloudFront delivery, or data transfer. Where two or three services drive most of the cost, a deep service-level rate beats a flat discount spread thinly across everything.
The analysis is straightforward. Rank your services by trailing twelve-month spend and by forecast growth. If the top three lines carry the majority of the bill, model a PPA rate card on those lines against the flat EDP discount on the whole. Run both against your forecast, not your history, because the rate card pays off most on the lines that grow. A PPA and an EDP are not mutually exclusive, and large buyers often run both: a baseline EDP discount across all consumption with service-specific rates layered on the concentrated lines.
Rate cards in a PPA are usually tiered against volume on the named service, and the tier definitions matter as much as the rates. A tier boundary set just above your realistic volume is a rate you will never reach, the same way a commitment floor set above your consumption is a discount you never collect. Negotiate tier breakpoints against your forecast volumes, and ask for the tier schedule as an exhibit to the agreement.
Action. Build a service-by-service spend distribution and bring it to the table; the account team already has it, and the buyer who arrives without it negotiates blind.
5Marketplace retirement and data transfer in the EDP
AWS Marketplace is the most underused instrument in EDP negotiations. A defined portion of eligible Marketplace purchases retires EDP commitment, which means third-party software you were going to buy anyway can count toward the floor. For a buyer carrying a large committed number, routing planned purchases of security tooling, observability platforms, or databases through Marketplace private offers converts an external cost into commitment progress.
| Discipline | What to do | Why it matters |
|---|---|---|
| Confirm eligibility | Get the eligible categories and retirement rate written into the agreement | Eligibility is contractual, not assumed |
| Route renewals | Move existing third-party contracts to Marketplace private offers as they come up | A private offer preserves negotiated pricing while adding retirement |
| Track quarterly | Reconcile retirement against the floor in the same review as consumption | A projected gap is visible while there is still time to act |
Private offers are the mechanism that makes the routing practical. A Marketplace private offer lets the third-party vendor present your negotiated price, payment schedule, and custom terms through the AWS Marketplace channel, so you keep the commercial terms you fought for while the transaction itself counts where you need it. Most major security, data, and observability vendors transact this way today, and the ones that resist usually have a channel-conflict reason that is theirs to solve, not yours to fund.
Data transfer belongs in the negotiation
Egress is the line buyers most often treat as fixed, and it is not. If your architecture moves significant data out of AWS to users, partners, or other clouds, the egress line grows with the business and is a fair target for a private rate inside the PPA or the EDP package. Put your egress forecast on the table alongside compute and storage. A buyer who prices data transfer at signature avoids re-opening the agreement when a data-heavy workload scales.
Action. Make Marketplace eligibility, the retirement rate, and an egress rate explicit terms of the deal. All three are contract language, and none survive as verbal assurances from an account team that rotates every two years.
6Shortfall, the overcommit trap, and exit flexibility
The overcommit is the most expensive EDP mistake, and it is built in at signature, not discovered at true-up. It happens when the committed floor, especially in the later ramped years, exceeds what the business actually consumes. The shortfall clause then converts the gap into an invoice: you pay for cloud you never used, at the moment the business is consuming less than planned. Protection is negotiated before signing or not at all.
| Risk | How it bites | The protection to negotiate |
|---|---|---|
| Ramp outruns adoption | A later-year floor exceeds real usage | Back-loaded ramp and a conservative first-year floor |
| Shortfall true-up | The gap to the floor is invoiced | Carry-forward of unused commitment into the next period |
| Business contraction | Divestiture or downturn cuts consumption | Reduction or exit rights for defined corporate events |
| Workload migration | A planned move off AWS strands the floor | Commit reduction tied to named workload events |
| Concentrated commitment | One stalled service drags the whole floor | Marketplace retirement and PPA spend counted to the commit |
Carry-forward is the protection most worth fighting for. The ability to roll unused commitment into the next period converts a hard penalty into a timing question, and it costs AWS little to grant when the relationship is growing. Exit and reduction rights for divestitures matter most for PE-owned businesses and any enterprise with an active portfolio strategy; name the events in the contract rather than relying on a renegotiation that will happen under pressure.
The governance cadence that prevents the shortfall
Shortfall is an operational failure before it is a contractual one. The agreement sets the floor, but the consumption that meets it is produced by dozens of engineering teams making independent decisions all year. Connect the two. Give FinOps explicit ownership of commitment tracking, set a quarterly review that compares the floor, the run rate, and the Marketplace retirement balance, and define in advance what happens when the projection shows a gap: which planned purchases accelerate into Marketplace, which workloads come forward, and at what point procurement opens a restructuring conversation with AWS rather than absorbing a true-up. Every expensive shortfall we have been brought in to remediate was visible at least two quarters before the true-up invoice arrived.
Action. Negotiate carry-forward and named exit events at signature, then stand up a quarterly FinOps-owned review so any projected gap surfaces while it is still cheap to close.
7Renewal: holding the commit to your own forecast
Renewal is where overcommitment compounds. The account team opens from your current committed floor plus growth, because committed spend growth is what they are measured on. A buyer who arrives without an independent forecast accepts that frame by default, and each renewal ratchets the floor further from reality.
Run the renewal as a fresh sizing exercise. Rebuild the bottom-up forecast from current workloads, planned migrations in both directions, and committed business change. Be prepared to hold the commit flat, or reduce it, when the data says so. A flat renewal from a position of evidence is a better deal than a grown renewal from a position of habit, and AWS accepts flat commitments from buyers who can show their maths far more readily than the first conversation suggests.
Start the renewal conversation at least six months before expiry, on the same timeline as the original deal. Bring the shortfall and retirement history from the closing term: if you consumed comfortably above the floor, that record argues for a better discount at the same floor, not for a higher floor. If you ran close to shortfall, that record is your argument for restructuring the ramp. Either way, your own consumption history is the strongest negotiating document in the room, and only one side brings it by default.
Action. Renew from your forecast, not from the vendor's growth assumption. The committed floor should be re-earned by the data at every renewal, never inherited from the last term.
8The term sheet review: what to verify before signature
Before signature, walk the term sheet against the checklist below. Every row is a clause we have seen cost a buyer money when it was assumed rather than written. The verification column is the question your legal and procurement team should be able to answer with a clause reference, not a recollection of what the account team said.
| Term | What to verify | Why it matters |
|---|---|---|
| Committed floor by year | The exact dollar floor for each period, including the ramp | This is the number you owe regardless of usage |
| Shortfall mechanics | How a gap is calculated, invoiced, and whether carry-forward applies | Converts a penalty into a timing question, or not |
| Marketplace eligibility | Which purchases retire commitment and at what rate | Assumed eligibility is the most common tracking error |
| Service exclusions | Which services or charge types sit outside the discount | Exclusions quietly shrink the effective percentage |
| Reduction and exit events | Named corporate events that permit commit reduction | Divestitures happen on business timelines, not contract ones |
| Renewal mechanics | Notice periods and what happens at expiry without renewal | Silence at expiry should not default you into worse terms |
None of this is adversarial. AWS negotiates these terms every day and respects buyers who do the same. The agreement you want is one both sides can administer without surprises, because the surprise clauses are the ones that surface three years later in front of your CFO.
Action. Require a clause reference for every line in the checklist before signature; anything answered from memory rather than the draft is not yet agreed.
Set the committed floor at the bottom of your high-confidence forecast, back-load the ramp, price your concentrated services on a PPA rate card, route eligible third-party spend through Marketplace, and write carry-forward, reduction rights, and named exit events into the agreement before you ever discuss the discount. The headline percentage is the last and least important number in the deal. The buyer who controls the floor, the ramp, and the protections recovers far more over a three-year term than the buyer who wins an extra point on a floor sized to the vendor's optimism.
Key takeaways
- The EDP discount is negotiable; the commitment floor is the real risk. Size it to spend you control.
- Sequence the twelve levers and negotiate the discount last, after structure is locked.
- Start six months out and spend the first half of that time on the bottom-up forecast.
- Spend above the floor still earns the discount, so set the floor at the bottom of your confidence range.
- Price concentrated services on a PPA rate card instead of accepting only a flat percentage.
- Make Marketplace eligibility, retirement rate, and egress pricing explicit contract terms.
- Negotiate carry-forward and named exit events, then track the floor quarterly through FinOps.
Frequently asked questions
What is an AWS EDP and how does the discount work?
The Enterprise Discount Program trades a committed minimum spend over a multi-year term for a cross-service discount on AWS usage. The percentage is private and scales with the size of the commitment, the length of the term, and the quality of your negotiation. Discounted spend still counts toward the commitment, so the discount and the floor interact in your favour when sizing is right.
What happens if we do not meet the EDP commitment?
You pay the shortfall. If actual usage falls below the committed floor for a period, AWS invoices the difference as a true-up. That clause makes conservative sizing and a back-loaded ramp more valuable than an extra point of discount, and it is the reason carry-forward language is worth negotiating.
Does AWS Marketplace spend count toward the EDP commitment?
A defined portion of eligible AWS Marketplace purchases retires EDP commitment. Routing planned third-party software purchases through Marketplace private offers helps you meet the floor with spend you were already going to make. Confirm eligibility and the retirement rate in your agreement before you rely on either.
Should we take an EDP or rely on Savings Plans and Reserved Instances?
Use both, layered in the right order. Savings Plans and Reserved Instances lower unit rates on specific resources, while the EDP discounts total spend in exchange for a commitment. Rate-discounted spend still retires the EDP floor, so the instruments stack: set the EDP commit first, then place Compute Savings Plans and Reserved Instances inside it by workload stability.
When is the best time to negotiate an AWS EDP?
Start six months before signature with a bottom-up forecast, and aim to close near an AWS quarter end or the December 31 fiscal year end, when timing pressure favours the buyer. Settle the commit size, ramp, and shortfall terms before you discuss the discount percentage.
Book a 30 minute call and get this playbook applied to your AWS agreement before you sign. Confidential, buyer side only.
Book a 30 minute callPrefer to start with the service detail? See how our cloud contract negotiation service runs an EDP engagement end to end, or return to the AWS EDP Negotiation Playbook overview.
Related research
Three companion guides extend this playbook across the rest of a cloud portfolio: the Azure MACC Negotiation Guide applies the same sizing discipline to Microsoft consumption commitments, the Cloud Renewal Strategy Playbook covers the renewal cycle across AWS, Azure, and GCP, and the Google Cloud Negotiation Guide maps the equivalent commitment mechanics on GCP.