BillSnap

Enable billing export to BigQuery

BillSnap reads from GCP's daily billing-export rows in BigQuery. If your project doesn't have an export configured yet, this is how to set it up. About three minutes.

You'll do this in 3 steps
  1. Open Cloud Billing → Billing export
  2. Click "Enable standard export" and pick a dataset
  3. Confirm "Enabled" and copy the dataset name

Total time: ~3 minutes for setup, then ~24 hours before the first row of billing data lands in the table.

Prerequisites: you need roles/billing.admin on the billing account, plus roles/bigquery.admin (or dataset-creation rights) on the GCP project where you want the export to live. Project owner is typically enough for both if it's your own project.

1 Open Cloud Billing → Billing export

Go to console.cloud.google.com/billing, select your billing account in the top-left dropdown, then click Billing export in the left sidebar (under "Cost management"). Make sure the BigQuery export tab is selected at the top.

Cloud Console Billing export page, BigQuery export tab. Standard usage cost is Disabled, with an 'Enable standard export' button.
Billing → Billing export → BigQuery export tab. The four exports show as Disabled; we only need the first one ("Standard usage cost").

Four exports are listed: Standard usage cost, Detailed usage cost, Pricing, and Committed Use Discounts. Only enable the first one.

Why only Standard usage cost? The detailed export has one row per resource (per VM, per disk) — typically 50-200× more rows than standard. BillSnap detects at the SKU level by design, so detailed export would inflate your BigQuery scan cost without changing what we can detect. Pricing and CUD exports are also unused by BillSnap.

2 Click "Enable standard export"

Under Standard usage cost, click the blue Enable standard export button. A dialog appears asking which BigQuery project and dataset should receive the export:

Edit cost dialog with a Projects dropdown set to 'My First Proj' and a Dataset field showing 'Select a dataset' error.
Edit cost dialog. Pick the BigQuery project that will hold the export, then either select an existing dataset or create a new one named billing_export.

Two fields to set:

  1. Projects. Pick which GCP project will host the BigQuery dataset that receives the export. Single-project setups: your one project. Multi-project orgs: a dedicated "billing" project. The export references costs across all projects under the billing account regardless of which project hosts the table.
  2. Dataset. Click the dropdown. If you have no existing datasets in that project, Google shows a yellow warning and offers "Create new dataset" inline. Name it billing_export (BillSnap defaults assume this name) or billing_export_ds — anything works, you'll tell BillSnap the exact name during onboarding.

Click Save.

Region note: when creating the dataset, pick a US or EU multi-region. BillSnap's reader handles both transparently. Avoid single-region zones unless you have a specific compliance reason.

3 Confirm "Enabled" and copy the dataset name

After saving, you're returned to the Billing export page. The Standard usage cost section now shows ✓ Enabled and the project + dataset name that will receive rows:

Billing export page after enabling. Standard usage cost shows Enabled (green check) with Project name and Dataset name: billing_export_ds. Detailed usage cost remains Disabled.
Standard usage cost is now Enabled. Project + dataset names are listed here — you'll paste them into BillSnap's connect form.

Note the Dataset name shown — in this example billing_export_ds. The export creates a table inside that dataset within a few minutes; the table name will be gcp_billing_export_v1_<BILLING_ACCOUNT_ID>. Confirm it exists via: 

bq ls --project_id=<your-project> billing_export_ds

      tableId                          Type    Time Partitioning
 -------------------------------------- ------- -------------------
  gcp_billing_export_v1_0086C9_E94769..  TABLE   DAY

Copy that full table name — you'll paste it into BillSnap's /connect form.

The first 24 hours are empty. The table is created right away, but Google's billing pipeline writes the first day's rows on a ~24-hour delay. New partitions land every morning UTC thereafter. Don't worry if bq ls shows the table but SELECT COUNT(*) returns zero on day one.

Continue to BillSnap onboarding

Once the table is created (~1 minute after enabling), you can finish connecting your project to BillSnap. You don't need to wait for the first day of data to land — BillSnap will start polling immediately, and the dashboard will show no anomalies for the first ~14 days until the baseline window fills.

Continue to /connect → See full walkthrough

Troubleshooting

"Permission denied" when clicking Enable standard export

You need roles/billing.admin on the billing account. Project owner is not enough — billing-account roles are separate from project roles. Ask whoever's listed as the billing-account admin to grant you the role or enable the export themselves.

I have multiple billing accounts — which one do I export?

The one that pays for the GCP project you want to monitor. Open the project's IAM & Admin → Settings page and look at "Billing account"; that's the one to configure.

"You have no existing datasets in this project"

That's the yellow warning shown in step 2 above. Just create one inline from the dropdown ("Create new dataset"), name it billing_export or billing_export_ds, pick a US or EU multi-region, and save. No need to leave the dialog.

The table exists but bq ls shows zero rows

Wait 24 hours. If after 48 hours there are still no rows, return to Billing export in the Cloud Console; the toggle may have reverted (it occasionally happens during org-policy changes).

Can I disable the export later?

Yes — click Disable standard export on the same page. The existing table stays in BigQuery (you can drop it manually if you want), but no new rows are written. BillSnap detection will halt within one daily cycle.