# BigQuery Data Quality

> Google Cloud's managed data quality scans (formerly Dataplex auto data quality) for BigQuery and Iceberg tables, with predefined and SQL rules, rule recommendations from data profile scans, and failed-record queries.

Source: https://toolsfordata.com/tools/bigquery-data-quality/ · Updated 2026-11-17

[Website](https://cloud.google.com/products/knowledge-catalog) · [Docs](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview)

Best for: Google Cloud teams that want scheduled data quality rules on BigQuery tables, managed inside their catalog.

## Overview

- Core Strength: BigQuery-native rules — scheduled scans, rules suggested from profiles, Terraform, lineage
- Test Types: Partial — SQL, table compare, failed rows
- Detection: Partial — profiling, pipeline monitoring
- Monitor Coverage: 4 of 11 types — as-code + UI
- Incident Workflow: Not confirmed
- Quality Analytics: Partial — health scores
- Lineage Depth: Full — table, column, end-to-end, impact analysis
- Governance Overlap: Partial — catalog, catalog sync, classification
- FinOps & AI Observability: Partial — cost monitoring
- AI Agent Features: Partial — auto test gen
- Developer Surface: Partial — API, Python SDK, MCP, scheduling
- Deployment Model: Partial — SaaS, cloud-native app
- Security & Access: Partial — SSO, RBAC, audit log, certifications
- Pricing: Usage — Knowledge Catalog processing
- Free Option: None
- Notable Caveat: Google Cloud only; email-only alerts and fixed thresholds, no anomaly detection

## What it's used for

BigQuery Data Quality is used to run scheduled data quality rules on BigQuery tables from Google Cloud's catalog: you start from rules recommended by a data profile scan.

## Pricing

BigQuery Data Quality has no separate license: Google Cloud bills the scans by the processing they use, under Knowledge Catalog (formerly Dataplex) pricing.

## Deployment

BigQuery Data Quality runs as managed scans in your Google Cloud project; there is nothing to install.

## Review

Strengths:
- Managed by Google Cloud: no servers or libraries to run.
- Rules can start from a profile scan's recommendations.
- Failed-record queries show which rows broke each rule.

Caveats:
- BigQuery and Iceberg tables only.
- Billed per scan, so frequent scans of large tables add up.

Recommendation: On Snowflake or Databricks, use Snowflake Data Quality or Databricks Data Quality. For checks across several warehouses, consider Soda Cloud or Monte Carlo; if your BigQuery models are in dbt, start with dbt tests.

## Pricing

- Model: platform usage
- Source: https://cloud.google.com/dataplex/pricing (checked 2026-11-10)

## Features

### Data Testing

- SQL Table Tests: Yes — Built-in rules check nulls, uniqueness, regex, accepted values, ranges and statistics, and SQL rules cover anything else. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- DataFrame Tests: No — Rules run on BigQuery and other catalog tables, not on in-memory DataFrames. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Unstructured Data Tests: No — No LLM or JSON Schema checks; the closest is regex on string columns. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Table Comparison Tests: Yes — A table-condition rule can compare aggregates such as COUNT(*) between two tables. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Data Diff: No — Cross-table rules compare aggregates, not rows. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Data Contracts: No — Rules are scan configurations; there is no producer-consumer contract. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Inspect Failed Rows: Yes — Each failed rule has a query in the console that returns the failing rows. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))

### Data Observability

- Data Profiling: Yes — Profile scans compute null percentage, distinct count, top values and summary statistics per column. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/data-profiling-overview))
- Anomaly Detection: No — Thresholds are fixed values in each rule; no learned thresholds are documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Freshness Monitors: No — No freshness monitor; freshness checks have to be written as SQL rules. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Volume Monitors: No — No volume monitor; row-count checks have to be written as SQL rules. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Schema Monitors: No — No alerts on column additions, removals or type changes are documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Field Monitors: Unknown — The profile Analysis tab trends column statistics across scans, but no deviation alerts are documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/data-profiling-overview))
- Distribution Monitors: Unknown — Profiles show top-value shares per scan, but no distribution drift detection is documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/data-profiling-overview))
- Segment Monitors: No — Row filters scope a scan to one segment; results are not broken down by a column. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Custom SQL Monitors: Yes — SQL rules run on every scheduled scan, and results are kept per job and can be exported to BigQuery. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Partition-Based Monitors: Yes — Incremental scans and row filters limit a scan to new rows or a partition. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- Window-Based Monitors: No — Scans cover the full table or new rows since the last scan; time windows are not documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- UI-Based Monitors: Yes — Scans and rules are created in a console wizard without code. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- Monitors-as-Code: Yes — Rules can be written in YAML for gcloud, and scans managed with Terraform from Git. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/manage-data-quality-rules-as-code))
- Pipeline Monitoring: Yes — Scan job history records state, start and end time and pass or fail per run. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))

### Incident Management

- Alert Notifications: Yes — Email reports to up to five addresses when the score drops below a target or a job fails; no Slack or Teams. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- Incident Triage: No — No incidents, owners or priorities. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Alert Feedback: No — No way to mark a failure as expected or snooze alerts. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Incident Resolution: No — No incident workflow. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Ticketing Integrations: No — No Jira, ServiceNow or Linear integration is documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- On-Call Integrations: Unknown — Cloud Monitoring can page PagerDuty, but no alert on data quality scan results is documented beyond email. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/monitoring))

### Data Quality Analytics

- Data Quality Dashboards: Unknown — Results show on each table's Data quality tab and in job history; trend dashboards need to be built in Looker from the BigQuery export. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Data Health Scores: Yes — Each scan reports a score, the percentage of rules passed, per table, column and dimension. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Incident Resolution Metrics: No — No MTTD or MTTR metrics. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))

### Data Lineage

- Table-Level Lineage: Yes — Google Cloud records lineage automatically for BigQuery copy, load, query and DDL jobs. ([source](https://docs.cloud.google.com/dataplex/docs/about-data-lineage))
- Column-Level Lineage: Yes — Column lineage is recorded for top-level BigQuery columns. ([source](https://docs.cloud.google.com/dataplex/docs/about-data-lineage))
- End-to-End Lineage: Yes — Lineage covers BigQuery, Dataflow, Spark, Airflow, Vertex AI and Looker, plus OpenLineage imports. ([source](https://docs.cloud.google.com/dataplex/docs/about-data-lineage))
- Impact Analysis: Yes — Lineage search follows downstream links to assess the impact of a change. ([source](https://docs.cloud.google.com/dataplex/docs/about-data-lineage))

### Data Governance

- Data Catalog: Yes — Knowledge Catalog is Google Cloud's searchable catalog, with a Data quality tab on each table. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Data Catalog Integration: Yes — Results publish to the table's entry in Knowledge Catalog; no push to third-party catalogs is documented. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- Data Classification: Yes — Sensitive Data Protection profiles BigQuery tables for PII and can send results to Knowledge Catalog. ([source](https://docs.cloud.google.com/sensitive-data-protection/docs/data-profiles))
- Data Products: Unknown — Knowledge Catalog has data products with SLAs and owners, but showing data quality per data product is not documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/overview))

### FinOps

- Query Performance: Unknown — BigQuery INFORMATION_SCHEMA.JOBS shows bytes and slot time per query, but data quality scans run in a Google project and don't appear there. ([source](https://docs.cloud.google.com/bigquery/docs/information-schema-jobs))
- Cost Monitoring: Yes — Scan costs carry labels in the Cloud Billing export, so cost can be attributed per scan and table. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/monitoring))

### LLMOps

- AI Observability: No — No monitoring of LLM or agent outputs. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/data-profiling-overview))

### Workflows

- Scheduling: Yes — Scans run hourly, daily, weekly, monthly, on a cron schedule or on demand. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- CI/CD Integrations: Unknown — Scans can be started from pipelines with gcloud, the API or an Airflow operator, but failing a pipeline on results is not documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/manage-data-quality-rules-as-code))
- Circuit Breakers: No — Auto data quality scans log results and send alerts or email reports on failures; no documented mechanism stops a pipeline on a failed scan. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Webhooks: Unknown — No HTTP callback on scan results is documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- REST / GraphQL API: Yes — A REST API creates, runs and reads data scans and generates rules. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- MCP Server: Yes — Google-hosted MCP servers for BigQuery and Knowledge Catalog let AI agents run SQL and search the catalog; there are no data quality tools of their own. ([source](https://docs.cloud.google.com/bigquery/docs/use-bigquery-mcp))
- Python SDK: Yes — The google-cloud-dataplex Python client creates and runs scans and reads results. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/use-data-profiling))
- dbt Integration: Unknown — No dbt integration is documented; the lineage API accepts OpenLineage events from any tool. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/open-lineage))

### AI Agents

- AI Chat Assistant: No — Data quality has no chat; Gemini in BigQuery generates table descriptions and queries. ([source](https://docs.cloud.google.com/bigquery/docs/data-insights))
- Natural Language Test Generation: No — Rules come from built-in types, SQL or profile recommendations; no plain-English rule creation. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- Automated Test Generation: Yes — Rules can be recommended from a profile scan's results, in the console or through the API. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- AI Issue Triaging: No — No AI root-cause analysis of failed rules is documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- AI Issue Resolution: No — No AI-proposed fix is documented. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))

### Security

- SSO: Yes — Google Cloud supports SAML and OIDC single sign-on through Workforce Identity Federation or Cloud Identity. ([source](https://docs.cloud.google.com/iam/docs/workforce-identity-federation))
- SCIM Provisioning: Unknown — Workforce Identity Federation SCIM covers only Gemini Enterprise and Looker; provisioning for the rest of Google Cloud goes through Cloud Identity. ([source](https://docs.cloud.google.com/iam/docs/workforce-identity-federation-scim))
- RBAC: Yes — IAM roles such as DataScan Administrator, Editor and Viewer control who can create and read scans. ([source](https://docs.cloud.google.com/dataplex/docs/use-auto-data-quality))
- Audit Log: Yes — Cloud Audit Logs record scan creation, changes, runs and reads of results. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/audit-logging))
- Compliance Certifications: Yes — Google Cloud holds SOC 2 and ISO 27001, among others. ([source](https://cloud.google.com/security/compliance/services-in-scope))

### Deployment

- SaaS Deployment: Yes — Scans run in Google-managed projects; nothing is installed. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Hybrid Deployment: No — Scans run only in Google-managed projects. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Self-Hosted Deployment: No — Not available as self-hosted software. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))
- Cloud-Native App: Yes — Scans run on Google-managed serverless compute and read BigQuery tables in place. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/auto-data-quality-overview))

### Open Source

- Open Source: No — A managed Google Cloud service; only a Terraform module is published. ([source](https://docs.cloud.google.com/knowledge-catalog/docs/manage-data-quality-rules-as-code))

## Integrations

4 integrations, each sourced from the vendor docs.

- Data Storage: Google Cloud Storage
- Data Warehouse: Google BigQuery
- Notifications: Email
- Open Table Formats: Apache Iceberg

## Alternatives

- [Snowflake Data Quality](https://toolsfordata.com/tools/snowflake-data-quality/): The Snowflake equivalent, with no extra vendor if your warehouse is Snowflake.
- [Databricks Data Quality](https://toolsfordata.com/tools/databricks-data-quality/): The Databricks equivalent, if your pipelines and tables live on Databricks.
- [AWS Glue Data Quality](https://toolsfordata.com/tools/aws-glue-data-quality/): The AWS equivalent, paid per use inside Glue.
- [Soda Cloud](https://toolsfordata.com/tools/soda-cloud/): Consider Soda Cloud for data contracts that cover BigQuery and the rest of your stack.
- [Monte Carlo](https://toolsfordata.com/tools/monte-carlo/): Consider Monte Carlo for ML monitors and lineage across BigQuery, BI and pipelines.
- [dbt tests](https://toolsfordata.com/tools/dbt-tests/): The free option if you transform with dbt: tests live next to the models.
