BigQuery

Refund Rules

BigQuery follows Google Cloud's standard policies - no refunds for consumed services (queries run, data stored), but customers can receive credits for documented service outages or billing errors. On-demand queries are charged per TB scanned after consumption, so no refunds for completed queries. For capacity pricing (slots), customers pay hourly for reserved compute regardless of usage, with no mid-cycle refunds.

Credit Policies

Google provides service credits via SLA guarantees - BigQuery offers 99.9% monthly uptime SLA with credits of 10-50% of monthly charges for verified downtime. Credits also available for: documented billing calculation errors, service performance issues below SLA thresholds, and proven data loss incidents. Credits are applied to future bills, not cash refunds.

Contractual Terms

Standard Google Cloud terms apply: month-to-month billing with no long-term commitments required for on-demand pricing. Capacity pricing (slots) requires 1-year or 3-year commitments for discounts. No automatic seat renewals since BigQuery uses consumption-based pricing. Price increases with 30-day notice. Customers can cancel/reduce usage anytime but committed slot reservations cannot be reduced until commitment expires.

Integration Complexity

Technical Notes

Simple REST API with excellent documentation. Authentication via Google Cloud service accounts or OAuth 2.0. Primary data sources: INFORMATION_SCHEMA views (queryable via BigQuery API), Cloud Billing API for cost data, Cloud Asset API for inventory. Key considerations: INFORMATION_SCHEMA is region-specific, 180-day retention limit for job history, requires bigquery.jobUser + bigquery.resourceViewer IAM roles. Rate limits are generous (100 queries/100 seconds per user). No webhooks - polling required for real-time monitoring.