The AWS EDP, benchmarks for an unpublished discount
An Enterprise Discount Program trades a multi year spend commitment for a percentage off nearly everything, and AWS publishes neither the tiers nor the terms. The observed range runs 5 to 20 percent, and the levers that decide your outcome are not the headline band: they are the commitment size, the ramp shape, and what counts toward retirement.
Prepared by Redress Compliance · August 6, 2026 · AWS advisory. Based on 20 to 25 EDP negotiations benchmarked 2024 to 2026.
Executive summary
The construction is simple, the tiers are secret. An EDP commits you to a minimum annual spend across a multi year term for a percentage discount, and observed discounts run from around 5 percent to around 20 percent, scaling with commitment size and term.
AWS publishes nothing, which makes benchmarks from comparable closed deals the only reference a buyer has, and the information asymmetry is the vendor's first advantage at the table.
The EDP stacks, it does not replace. The discount applies on top of Savings Plans and Reserved Instances, not instead of them, so the honest baseline for any EDP model is your post SP and RI effective rate.
Deals modeled against on demand pricing flatter the EDP by exactly the optimization already done, the same double counting error every commitment construct invites.
Retirement scope is a first order lever.
Whether AWS Marketplace spend counts toward the commitment changes the math materially: in our negotiations, getting marketplace spend included raised commitment retirement by 10 to 25 percent.
Converting third party software bills into commitment progress and shrinking the organic growth the commit actually requires.
Right sizing beats band chasing. A missed commitment is usually billed as shortfall, and 2 of 5 customers who chased the top discount band over committed, paying shortfall or carrying unused commitment that erased the extra points they chased.
A back loaded ramp, small early years, growth weighted late, cut early year shortfall risk to near zero in most engagements, at the cost of nothing but asking.
How the EDP works, and what it stacks with
| Element | How it behaves | The buyer side note |
|---|---|---|
| The commitment | A minimum annual spend, per year, across a multi year term | Sized to your defensible floor, never to the vendor growth model |
| The discount | A percentage off most AWS spend, observed at 5 to 20 percent | Applies on top of Savings Plans and RIs; model against the post optimization rate |
| The ramp | Annual commitments can step up across the term | Back loaded ramps match real growth and neutralize early shortfall risk |
| The shortfall | A missed annual commitment is typically billed anyway | The asymmetry that makes over commitment the expensive direction |
| Marketplace scope | Third party purchases may or may not retire commitment | The 10 to 25 percent lever, negotiated into the definition of qualifying spend |
The discount is public math on a private menu. Because AWS publishes no tiers, every EDP conversation starts from what the account team offers, and the only counterweight is knowing what comparable commitments actually signed.
The benchmark is not decoration on the negotiation; on an unpublished program, it is the negotiation.
Sizing the commitment, the floor and the ramp
The sizing discipline is the same one every commitment construct in this library demands: twelve months of spend, the post Savings Plan effective baseline, known migrations and decommissions subtracted, and the commitment set at the defensible floor rather than the growth story.
The EDP adds the ramp as its own lever: annual commitments that step up across the term let the early years match reality while the discount prices against the full term commitment.
The band chasing failure is worth naming precisely because it feels like negotiating: stretching the commitment to reach the next discount tier trades a certain liability for a conditional saving, and in our file 2 of 5 who did it lost the trade.
Five points more discount on spend you achieve is worth less than zero shortfall on spend you do not, and the arithmetic only reads that way in advance.
The AWS EDP negotiation playbook
The commitment sizing worksheet, the ramp structures, the marketplace inclusion language, and the benchmark bands by commitment size and term.
Get the white paper →The marketplace lever, retiring commitment with software you already buy
Marketplace inclusion converts third party software purchases, the Databricks, Confluent, and security tooling bills you pay anyway, into EDP commitment retirement.
In our negotiations it raised retirement by 10 to 25 percent, which either shrinks the organic AWS growth the commitment requires or supports a larger commitment at the same risk.
The route interacts with the software negotiations themselves, as the Confluent commit guide works from the vendor side: marketplace routing changes whose discount applies, so the two negotiations belong in one model.
- Percentile standing for your exact deal size and industry, from real closed transactions
- Scenario simulation before the call: test alternative terms and see the financial impact of each
- A negotiation playbook, talking points, and a two page executive brief on day one
What we saw across EDP negotiations, 2024 to 2026
Across roughly 20 to 25 AWS EDP negotiations Redress benchmarked between 2024 and 2026, the customers who chased the highest discount band were the ones most likely to over commit:
Over committed for the top tier, then paid shortfall or carried unused commitment that erased the extra points.
Commitment retired by third party software spend once inclusion was negotiated into the qualifying definition.
The structural winner was boring: right sized commitments on back loaded ramps, marketplace included, renewed from measured consumption rather than renegotiated from aspiration.
The same floor discipline runs through the Google Cloud CUD guide and every commitment construct on the estate, because the vendor changes and the asymmetry does not.
Your first five moves
- Baseline the post optimization rate: spend after Savings Plans and RIs, by workload, over twelve months. The EDP models against this, never on demand.
- Commit to the defensible floor with known changes subtracted, and refuse the tier stretch: shortfall risk prices higher than discount points.
- Shape the ramp back loaded, small early years and growth weighted late, so the commitment meets reality on its way up.
- Negotiate marketplace inclusion into the qualifying spend definition, and route the third party purchases that benefit through it deliberately.
- Bring benchmarks to an unpublished program: comparable commitments and their signed bands are the only market data that exists. The AWS practice brings them with you, on your side of the table.
Frequently asked questions
What discount does an AWS EDP give?
Observed discounts run from around 5 percent to around 20 percent off most AWS spend, scaling with commitment size and term, in exchange for a multi year minimum annual spend commitment. AWS publishes no tiers, so benchmarks from comparable closed deals are the only reference available to a buyer.
Does the EDP replace Savings Plans and Reserved Instances?
No, it stacks on top of them, which means the honest model prices the EDP against your post Savings Plan effective rate rather than on demand pricing. Deals modeled against on demand double count the optimization already done and overstate the EDP's value by exactly that amount.
What happens if we miss our EDP commitment?
A missed annual commitment is usually billed as shortfall: you pay the committed amount regardless of consumption. That asymmetry makes over commitment the expensive direction, and it is why right sizing to the defensible floor consistently beat chasing the top discount band in our negotiations.
Does AWS Marketplace spend count toward the EDP?
It depends on what the agreement's qualifying spend definition says, which makes it a negotiation lever rather than a fact. Getting marketplace purchases included raised commitment retirement by 10 to 25 percent in our file, converting software bills you pay anyway into commitment progress.
Should we increase the commitment to reach a higher discount band?
Rarely. Stretching trades a certain liability for a conditional saving, and 2 of 5 customers who chased the top band in our negotiations over committed, paying shortfall or carrying unused commitment that consumed the extra points.
Extra discount on achieved spend is worth less than zero shortfall on unachieved spend.
What is a back loaded EDP ramp and why ask for one?
Annual commitments that start small and step up across the term, matching real growth instead of averaging it. In most of our engagements the back loaded shape cut early year shortfall risk to near zero, and it costs nothing except knowing to ask during structuring rather than after signature.
