AWS usage export

This guide walks you through creating a recurring AWS Data Export (Cost and Usage Report 2.0) delivered to an Amazon S3 bucket you own. Create the export from the management (payer) account so it covers every member account in one export.

Before you start

You need:

• Access to the management (payer) account with permission to create Data Exports, so the report covers all member accounts.
• An Amazon S3 bucket to receive the export. Create one in this account, or use an existing bucket, and note the name and prefix. If you plan to use Greenpixie's cloud usage report enrichment service, this is the same bucket you will connect later, so a name that includes your company or team is recommended, for example abc-corp-cur-data. See Connect Amazon S3.
• A bucket policy that allows delivery by the export service principals billingreports.amazonaws.com and bcm-data-exports.amazonaws.com. The console offers to apply this during creation. If bucket policies are managed through IaC, apply it through the pipeline first. See the AWS S3 setup guide for data exports.

Two conditions to check:

• If the bucket uses SSE-KMS with a customer-managed key, the key policy must also allow those service principals, or deliveries will not arrive.
• If a CUR or data export already exists for other tooling, create a second export with the settings below. Do not repoint the existing one.

If usage spans more than one AWS organisation, repeat these steps per management (payer) account and use a distinct prefix per organisation.

Create the export

1. Sign in to the management (payer) account and open the Billing and Cost Management console.
2. Go to Data Exports and choose Create export.
3. Choose Standard data export. Name it greenpixie-cur2, or follow your naming convention.
4. Select the Cost and Usage Report 2.0 (CUR 2.0) table.
5. Under Column selection, keep all columns, the default full set.
6. Set the delivery destination to s3://<destination-s3-bucket>/<prefix>. Accept the bucket policy update when prompted, unless it is applied via IaC.
7. Create the export.

Export settings

Cost allocation tag columns are optional. Include them if you want output attributable to teams and applications.

Required directory structure

If you will be using Greenpixie's cloud usage report enrichment service, the delivery path must follow the shape below.

<bucket_name>/<input|output>/<platform>/<account_id>/<year_month>/{partition files}

s3://<destination-s3-bucket>/input/aws/<payer_account_id>/2026-07/...

<bucket_name>/<input|output>/<org_id>/<platform>/<account_id>/<timestamp>/{partition files}

s3://<destination-s3-bucket>/input/org_12345/aws/<payer_account_id>/20260722T120000Z/...

Data and format notes

Export all columns and keep the full default set. The complete set gives you the most context for later analysis.

Accepted alternatives to CUR 2.0:

• FOCUS-format exports. Select the FOCUS with AWS columns table instead of CUR 2.0, with the same destination and settings, and do not create both. AWS offers both a "FOCUS 1.0 with AWS columns" table and a "FOCUS 1.2 with AWS columns" table.
• Legacy CUR, if CUR 2.0 is not available.

If you will be using Greenpixie's cloud usage report enrichment service, Parquet with zstd compression is preferred (the console's default Parquet compression, or CSV with gzip, also work), and no single file should exceed 500 MB. CUR 2.0 partitions large deliveries automatically, though this is not a hard 500 MB cap.

Good to know

• The first files can take up to a day to arrive, and AWS then refreshes the export up to once per day.
• The export delivers data from creation onward. If you need historical months for a baseline, raise the CUR 2.0 backfill options with the AWS account team when creating the export.
• At hourly granularity with resource IDs, files can be very large in a big estate. This is expected. Apply bucket retention per policy once the cadence is confirmed.

Check it worked

You should have:

• The export showing as active in Data Exports.
• Files landing under the agreed prefix in <destination-s3-bucket> within 24 hours of creation.
• Line items at hourly granularity, including resource IDs.
• Parquet files, with zstd preferred, and no single file larger than 500 MB.

To remove the export later, disable or delete it in Data Exports. Nothing else depends on it.

Next steps

Once your export is running, you can enrich it with energy, carbon and water metrics in two ways.