For Data owners · Updated Nov 6, 2026

Share data through Google Cloud Storage

For teams already on Google Cloud, your team writes scheduled exports to a Cloud Storage bucket you own and grants Scowtt read-only access. Your data stays in your project, you control retention, and you can revoke access at any time. Scowtt never needs write access, keys or passwords.

How it works: Scowtt sends you the service account to grant. You create a bucket, give that account read-only access, schedule exports and send Scowtt the details. Scowtt checks the data with you.

Before you start

A Google Cloud project with billing enabled.
Someone who can create buckets and manage bucket access, for example with the Storage Admin role.
The Scowtt service account (or accounts) to grant, from your Scowtt team. The main one looks like scowtt-user@yourcompany-scowtt.iam.gserviceaccount.com.

Steps

You: Tell your Scowtt team you'll use Cloud Storage

Let your Scowtt team know which tables you'll export and how often.
Scowtt: Send the service account to grant

Your Scowtt team sends the Scowtt service account (or accounts) to grant. The main one looks like scowtt-user@yourcompany-scowtt.iam.gserviceaccount.com.
You: Create a bucket

Create it in a US location with uniform bucket-level access turned on, for example yourcompany-scowtt-export. A dedicated folder in an existing bucket also works.
```
gcloud storage buckets create gs://yourcompany-scowtt-export --location=US --uniform-bucket-level-access
```
You: Grant each service account two roles

Grant Storage Object Viewer and Storage Legacy Bucket Reader. In the console, open the bucket, then Permissions > Grant access. From the command line, run this once per account and role:
```
gcloud storage buckets add-iam-policy-binding gs://yourcompany-scowtt-export --member=serviceAccount:<scowtt service account> --role=roles/storage.objectViewer
```
Then repeat with --role=roles/storage.legacyBucketReader. Without it, Scowtt's reads fail with a storage.buckets.get error.

Google's step-by-step guide: Add a principal to a bucket-level policy.

You: Organize your files

Use one folder per table, partitioned by date and hour:

gs://yourcompany-scowtt-export/<table>/dt=YYYY-MM-DD/hour=HH/part-00000.parquet

You: Load the backfill, then schedule exports

Load the historical backfill first, then schedule recurring exports, ideally hourly. Most warehouses export to Cloud Storage natively; see Export data from Snowflake or Export data from Amplitude.
You: Send Scowtt the bucket details

The bucket and folder path, the tables included, the file format, the backfill date range, the refresh schedule and a technical contact.
Scowtt: Check the data with you

Scowtt confirms it can read the bucket, checks row counts, schema and sample records with you, then confirms the recurring schedule.

No bucket of your own?

If your team doesn't run Google Cloud, or your tool writes files out (Snowflake, Adobe Analytics Data Feeds and similar), Scowtt can provide the bucket instead. See Send files to a Scowtt-managed bucket.

Good to know

Policy error when granting access? Your organization restricts sharing to its own domain. Ask your Scowtt team for Scowtt's Google Workspace customer ID. Your Google Cloud admin can allow it in that policy, or add an exception for this bucket.
Exporting from BigQuery? The bucket must be in the same location as the dataset. If your data is already in BigQuery, a direct share is less work; see Share data through BigQuery.
Agree a retention period with your Scowtt team before setting rules that delete old files.
Follow the data format standards for formats, folders and file names.
To end access, remove the Scowtt service account from the bucket.

Who does what

Step	Who	What changes hands
Tell your Scowtt team you'll use Cloud Storage	You	Which tables, how often
Send the service account to grant	Scowtt	Service account name(s)
Create a bucket	You	—
Grant each service account two roles	You	Read-only access for the Scowtt service account
Organize your files	You	—
Load the backfill, then schedule exports	You	History files, then recurring exports
Send Scowtt the bucket details	You	Bucket and folder path, tables, format, date range, schedule, contact
Check the data with you	Scowtt	Confirmation, or fixes needed

Share data through Amazon S3You keep full ownership of your S3 bucket. Your team creates a role in your own AWS account that gives Scowtt read-only access to a specific bucket or folder. No access keys or long-lived credentials are ever shared, and you can revoke access entirely on your end at any time.Data format standardsFor file-based delivery through S3, GCS or SFTP, exports should follow these standards.Choose how to connect your dataThere are several ways to share CRM and analytics data with Scowtt. Your Scowtt team will help you pick the one that fits your environment best.