The AWS Private Pricing Agreement in 2026. How to test the custom rates before you sign.

An AWS Private Pricing Agreement can beat the standard commitment discount on the services you use most. Without an independent benchmark, though, you cannot tell whether the custom rate is competitive or just below list.

How does an AWS Private Pricing Agreement work in 2026?

An AWS Private Pricing Agreement (PPA) sets custom negotiated rates on selected AWS services, on top of the standard commitment discount, with every term agreed privately. AWS publishes its general pricing model on the AWS pricing page, and a PPA rate is a private departure from it. The published list alone cannot tell you whether your offer is good.

Custom rates on the services that dominate your bill

A PPA rate is worth most on the services that make up the bulk of your invoice. A 20 percent cut on a service that costs you $40,000 a year changes little. The same cut on a $3 million storage or data transfer line changes the whole deal. You can see where your spend concentrates in AWS Cost Management.

• Targeted rates. Custom pricing applies to the services and usage types named in the agreement, usually your highest spend lines.
• Linked to a commitment. AWS usually ties the rates to a usage or spend commitment, sometimes on that specific service.
• Fixed to the term. The rates hold for the agreement term, which cuts both ways when public prices change.

How the custom rate reaches your invoice

PPA rates are applied as a billing adjustment against eligible usage, so your billing data shows the private rate as a discount line against the usage. Third party software bought through AWS Marketplace private offers is negotiated privately with each seller, and eligible products draw down the same AWS commitment. Price both in the same round.

How does a PPA differ from an Enterprise Discount Program?

An EDP gives a broad percentage discount across eligible spend in return for a commitment. A PPA rate cuts the unit price on specific services. The EDP rewards total spend and the PPA rewards concentration, so they solve different problems and often work best together in the same agreement.

When service rates add value beyond the commitment discount

• Concentrated spend. If three or four services make up most of your bill, a custom rate on them beats a flat percentage spread across everything.
• Steady, large usage. Storage, data transfer and content delivery at high, predictable volume are the classic candidates, because AWS can price them with confidence.
• Layered discounts. For the right profile, the broad commitment discount plus service rates beats either one alone.

A worked example: why the two parts must be modeled together

Lower unit rates reduce the spend that counts toward your commitment. That is harmless while you have headroom. Say a company would spend $10.0 million a year at list, $3.0 million of it on S3 storage and data transfer, and commits $8.5 million a year, measured on spend after discounts, at a hypothetical broad discount of 10 percent.

AWS then offers a custom rate 20 percent below list on the storage and transfer lines, replacing the broad discount there. The table shows the net bill with and without that rate, first at planned usage and then if usage falls 10 percent.

At planned usage the service rates save $300,000 a year. If usage drops 10 percent, both versions cost exactly the commitment, because the lower rates only deepened the shortfall. The service rate pays off only while your net spend stays above the commitment.

So either lower the commitment when you add service rates, or ask for PPA services to count toward it at their value before the custom discount. Third party software handled through AWS Marketplace belongs in the same model, since eligible Marketplace spend can fill part of a gap.

What have we seen in recent AWS PPA negotiations?

The finding that recurred most was a custom rate accepted without an independent benchmark to test it. That held across roughly 20 to 30 AWS Private Pricing Agreement negotiations I advised on between 2024 and 2025, and it showed up in three patterns.

• Unbenchmarked rates. Where nothing but list price sat behind the negotiation, 8 to 18 percent of the achievable rate reduction was left unclaimed.
• Rates on the wrong services. Custom discounts landed on services where usage was light, while the services that drove the bill stayed on the broad discount.
• No review clause. The agreements had no rate review rights, so buyers kept paying the old rates while AWS list prices fell during the term.

Why a rate below list is a weak test of a good deal

The usual advice is that any custom rate below list is a good deal, worth signing before the account team changes its mind. We disagree. In more than half the PPAs we reviewed in 2024 and 2025, the rate looked attractive against list but sat above what a benchmarked negotiation achieved, by a margin in that same range.

Without a review clause, the gap then grew each time list prices fell. List is the one price AWS sets for customers who negotiate nothing. Benchmark every PPA rate against what comparable buyers pay for the same service at similar volume, aim the custom rates where your spend concentrates, and write in rate review rights before you sign.

How should you structure the PPA rates?

Build the PPA around your spend concentration and a commitment you are confident of meeting. Custom rates go on the services that dominate the bill, the commitment sits at or below your trailing usage, and the contract says what happens to your rates when AWS prices move.

Contract terms to ask for

• Rates stated as a discount from the public price. A percentage follows list down automatically when AWS cuts a price. A fixed unit rate does not, and you need a separate clause to capture the cut.
• An annual rate review. A right to reopen the service rates each contract year, measured against current list and your actual volume.
• Protection against list increases. AWS also raises some prices. It updates EC2 Capacity Blocks for ML prices on a published schedule, and those rates rose during 2026. Ask that your agreed rates hold if list rises on a covered service. Our note on the Capacity Blocks price protection gap covers the wording.
• A clear definition of eligible spend. State whether PPA services count toward the commitment before or after the custom discount, and how Marketplace purchases count. The commitment measurement clause is where this gets settled.
• Service substitution rights. If you move a workload from one service to another, the commitment and the rate should move with it. See our guide to substitution and swap rights.

For clause by clause redline wording, see our PPA clause redlines.

How the right structure changes with your spend profile

A company spending around $2 million a year across a dozen services usually gets more from a good broad commitment discount and clean Savings Plans coverage than from service rates. There is rarely enough volume on one line for AWS to price it privately. Compute rates also interact with Savings Plans, which our effective rate analysis works through.

At $20 million or more, with storage, transfer or a managed database making up a large share, service rates become the bigger part of the value. Here the benchmark and the review clause matter most, because one percentage point on a hypothetical $6 million storage line is $60,000 a year, for every year of the term.

How do you benchmark and prepare for an AWS PPA negotiation?

Prepare three things before the first pricing call: a spend concentration analysis, an independent rate benchmark for your top services, and a commitment sized to what you actually use. In our negotiations AWS moved custom rates when the buyer brought a credible comparison, and rarely moved them when the only reference was list.

Reports that show where your spend concentrates

• AWS Cost Explorer. Pull 12 months at monthly granularity, grouped by service, then by usage type within your top five services. Use the net unblended cost view to see what you pay after existing discounts.
• AWS Data Exports. In the Cost and Usage Report, the line items with the Discount type show each discount AWS applied, and the net unblended cost column shows the actual cost after discounts.
• AWS Pricing Calculator. Price your top usage types at current list, so every benchmark comparison uses the same volume and region.
• Marketplace invoices. List the third party software you already buy, and which of it could move to private offers that draw down the commitment.

Then compare your current and offered rates for those usage types against independent market data for similar volume. Our AWS discount benchmarks show what the comparison looks like, and this note covers how to use a benchmark in the room without handing it to AWS.

What the account team will say, and what to say back

Mistakes that cost buyers the most

1. Negotiating the service rates and the commitment in different meetings. As the worked example shows, a better rate can raise your shortfall risk if the commitment does not move with it.
2. Accepting a service rate on a line you plan to shrink. If a migration or cleanup project will cut that usage, the rate protects little and the service commitment becomes a liability.
3. Comparing offers on credits. A large credit in year one can hide a weak rate that you pay for the whole term.
4. Letting the benchmark age. A rate comparison from the previous renewal misses every price change since.

On an AWS PPA a rate below list feels like a win, but only a benchmark tells you whether it is the rate you should have signed.

Our AWS EDP negotiation guide covers sizing the commitment itself, and our white paper on AWS Marketplace procurement strategy and Channel Partner Private Offers covers the third party side of the same agreement.

What to do next

1. Find the concentration. Export 12 months of spend by service and usage type from AWS Cost Management and identify the few lines that drive the bill.
2. Benchmark those rates. Compare your current and offered rates on the highest spend services against independent market data for your volume and region.
3. Target the custom rates. Ask for PPA rates on the concentrated services and decline rates on lines you use lightly or plan to shrink.
4. Size the commitment to a confident baseline. Base it on trailing 12 month usage, and lower it when service rates reduce the spend that counts.
5. Model the rates and the commitment together. Run planned and lower usage cases so a better rate does not turn into a larger shortfall.
6. Write in review rights. Get an annual rate review or a percentage off public price, plus protection if list rises on a covered service.
7. Take the evidence into the room. Open the negotiation with the benchmark and the concentration analysis, so the discussion starts from market rates.

Frequently asked questions