Data Governance Tools: The Best Data Governance Software and Platforms for 2026

Collibra is the reference implementation of the category: business glossaries, policy management, stewardship workflows, data quality, privacy, and the reporting an auditor expects. It markets a Leader placement in the 2026 Gartner Magic Quadrant for Data and Analytics Governance Platforms. Two operational details buyers miss until late in the cycle. Column-level lineage is real but is not generated for tables created by SQL statements unless you supply the SQL through the folder-connection method, and the CLI lineage harvester reached end of life on 31 July 2026, so connectivity now runs through Edge sites. Collibra also acquired Raito in June 2025 and Deasy Labs in July 2025, and its 2026 positioning is an AI control plane rather than a catalog.

Salesforce completed its acquisition of Informatica on 18 November 2025 in a deal valued at roughly 8 billion dollars, and the governance product survived intact as Cloud Data Governance and Catalog inside IDMC, now branded Informatica from Salesforce. It was named a Leader in the January 2026 Gartner Magic Quadrant for Data and Analytics Governance Platforms. It is the strongest option if you also need master data management in the same contract, and the widest in scope of anything here. Pricing runs on consumption units called IPUs; the model is documented but Informatica publishes no dollar rate per IPU. Hybrid and on-premises sources need a customer-installed Secure Agent sized at 16GB of memory minimum.

Alation approaches governance through discovery and adoption rather than policy: it analyzes query logs behaviorally to surface which tables people actually trust, and its search experience is genuinely better than most of this list. In July 2026 it relaunched as AIOS, an intelligence operating system for enterprise AI, after acquiring Numbers Station. The detail to pin down in a contract is lineage granularity. Alation documentation states that table-level lineage is the default and that column-level lineage requires per-connector enablement, and for Snowflake specifically it requires a paid parser add-on. If column-level lineage is a governance requirement rather than a nice-to-have, get it quoted explicitly instead of assuming it is bundled.

Atlan is the best of the modern platforms on time to value, and it markets a Leader placement in the 2026 Gartner Magic Quadrant for Data and Analytics Governance Platforms. Column-level lineage is core rather than an add-on, backed by dedicated column-process entity types and native OpenLineage ingestion from Airflow, Spark, and dbt Cloud. Governance here means tags, classifications, ownership, and policy propagation across the graph rather than the heavy approval chains Collibra and Informatica ship. Architecturally it is a single-tenant SaaS control plane in a dedicated VPC plus a Kubernetes Secure Agent running inside your environment, so it is not self-hostable end to end. Pricing is a conversation, not a page.

Purview is the only enterprise option on this page that publishes exact rates, which makes it the easiest to budget before a sales call. Unified Catalog bills per unique governed asset per day plus data governance processing units for quality and health jobs, and both rates are queryable from the Azure retail price API. The trap is naming: three different things carry the Purview label. Azure Data Catalog was retired on 15 May 2024, the classic Purview Data Catalog is described by Microsoft as no longer taking on new customers and in customer support mode with no published end-of-life date, and Unified Catalog is the current product. Column-level lineage is supported, but manual column-level lineage is unsupported when a process asset sits in between, and automated lineage only arrives from a fixed list of process systems.

Ataccama ONE bundles governance, data quality, master data management, and observability in a single platform, and its center of gravity is genuinely data quality: it markets a Leader placement in the 2026 Gartner Magic Quadrant for Augmented Data Quality Solutions. That makes it a better fit than a catalog-led tool when the governance mandate came out of a regulatory finding about incorrect data rather than out of a discovery problem. It runs as SaaS or self-hosted, which matters for organizations that cannot ship metadata out. Ataccama publishes no pricing. Third-party figures circulate for it, including a widely repeated annual number that does not come from Ataccama, and we do not reprint them.

Qlik acquired Talend in May 2023 and now sells the governance capability as Qlik Talend Cloud and Talend Data Fabric, combining data integration, data quality, stewardship, and cataloging in one stack. Qlik markets a Leader placement in the 2026 Gartner Magic Quadrant for Augmented Data Quality Solutions. It is the right shape when the same team owns the pipelines and the governance obligation, because policies and quality rules sit next to the transformations. The important correction, since guides still recommend it as a free option: Talend Open Studio, the open-source edition, reached end of life at version 7.3 on 31 January 2024, downloads were removed, and the old product page now redirects to a commercial Qlik page. There is no free Talend to evaluate.

erwin has been part of Quest Software since 2021 and remains one of the more underrated options, particularly where erwin Data Modeler is already in place and the metadata can be harvested from models rather than reverse-engineered from a warehouse. Quest shipped erwin Data Intelligence 15 with automated AI model certification, a per-asset trust score built from nine components including quality, governance completeness, timeliness, and popularity, and more than 100 first-party connectors with automated lineage. It is a serious enterprise product with a smaller install base and correspondingly less community material, so plan for vendor-led enablement rather than finding answers on a forum.

Immuta solves a different half of the problem and is often bought alongside a catalog rather than instead of one. It automates discovery and classification, then enforces access through policy-as-code, dynamic masking, attribute-based access control, and privacy-enhancing techniques, pushing enforcement natively into Snowflake, Databricks, BigQuery, and Starburst rather than proxying queries. In June 2026 it expanded its Databricks partnership with agentic data access, intent-driven access control built on Unity Catalog role-based access controls, a Comply app for Unity Catalog, and a scalable attribute-based architecture for group permissions. If your governance mandate is who may see which rows, this is the category to shortlist. It will not tell you what a column means or where it came from.

data.world was acquired by ServiceNow, announced in May 2025 and closed that July, and now also appears as Data Catalog inside ServiceNow Workflow Data Fabric. Its distinguishing technical choice is a genuine knowledge graph underneath, which makes relationship-heavy governance questions easier to express than in a relational metadata store. The strategic read is straightforward: if your organization runs its work in ServiceNow, putting governance requests into the same workflow engine that already handles change management removes a real adoption barrier. If you do not run ServiceNow, weigh how a standalone product is resourced inside a platform company with a very different core business.

Secoda joined Atlassian in December 2025 and is heading toward Rovo and the Atlassian System of Work. It has always been the least ceremonious tool in this space: fast to set up, AI-assisted documentation, and a deliberate bias toward getting analysts to use it rather than toward satisfying an audit. That makes it a good fit for a team of five to fifty people who want ownership, documentation, and lineage without standing up a governance council, and a poor fit for a regulated program that needs attestation trails. As with every acquisition on this page, ask how the standalone roadmap is funded and what renewal looks like when the parent wants a bundle.

OpenMetadata is Apache-2.0 licensed and, unusually for open source, genuinely covers governance rather than just cataloging: glossaries, classification, ownership, policies, data quality tests, and real column-level lineage. Collate, the company behind it, publishes tier limits including a free tier at five users and 500 assets but no dollar figures. The honest cost is operational. A production deployment wants a Java server, a relational database, a search cluster, and an ingestion runner, roughly 10 vCPU and 40GiB of memory before anyone logs in, breaking changes have shipped in patch releases, and Snowflake lineage ingestion needs privileges at the ACCOUNTADMIN level. Airflow has not been mandatory since v1.12.

Acryl Data renamed itself to DataHub on 21 May 2025 alongside a 35 million dollar Series B, and the open-source project is DataHub Core under Apache-2.0. Column-level lineage is in the free edition rather than gated behind the paid cloud, currently across 53 of roughly 103 connectors, with Kafka, object storage, Delta Lake, and most BI-adjacent sources remaining table-level. The project is healthy: v1.6.0 in May 2026, more than 100 commits a month, over twelve thousand stars. Governance features are real but more metadata-centric than workflow-centric. Infrastructure is the cost: Kafka, a relational database, a search index, a graph index, and four services, with the Kubernetes guide asking for more than 7GB of memory. DataHub Cloud publishes no pricing and its pricing URL returns a 404.

Atlas is the original open-source governance project and is still the right answer for one specific situation: an existing Hadoop estate running Hive, Spark, and Kafka where Atlas hooks are already wired in. Its classification and tag-based policy model, particularly paired with Ranger for enforcement, remains coherent. For a modern cloud warehouse it is the wrong tool. Connectivity to Snowflake, BigQuery, and Databricks is community-maintained rather than first-party, the operational footprint assumes a Hadoop-era stack, and the user interface reflects its age. If you are choosing new infrastructure in 2026, evaluate OpenMetadata or DataHub instead, and keep Atlas on the list only if it is already running.

Databricks Unity Catalog

The most complete platform-native governance layer. Unity Catalog handles fine-grained access control, audit, and automatic column-level lineage for queries run on Databricks, aggregated across every workspace attached to the metastore, and it is on by default for workspaces created after November 2023. Two corrections, because comparison pages repeat both errors: lineage in Catalog Explorer is retained indefinitely rather than for 90 days, with system tables holding a rolling one-year window, and the long list of exclusions those pages cite has largely been removed from the documentation. Real remaining limits are nothing before 1 September 2024, no lineage preserved through a rename of a catalog, schema, table, view, or column, and no column lineage when the source or target is referenced by path rather than table name. No separate charge, but it requires the Premium plan or above.

Snowflake Horizon Catalog

Horizon is an umbrella brand over individually documented governance features rather than a product you install: tagging, classification, masking and row access policies, data metric functions, the Trust Center, and lineage. Column-level lineage went generally available in release 9.3 in February 2025, requires Enterprise Edition or higher, retains one year of history, and covers nothing before November 2024. Process lineage for procedures and tasks went generally available in October 2025. Horizon is not a billable line item, though components such as data quality monitoring and classification meter serverless compute. Keep the names apart, because they get conflated constantly: Horizon is the governance catalog, Open Catalog is the managed Apache Polaris service for Iceberg and is now closed to new customers, and Openflow is NiFi-based ingestion and not governance at all.

Google Knowledge Catalog

Renamed from Dataplex Universal Catalog on 10 April 2026, with API, CLI, and IAM names unchanged and the pricing SKUs still carrying the Dataplex name. The legacy Data Catalog service began a phased shutdown on 1 June 2026. Governance covers policy tags, column-level access control, data profiling, and quality scans. Column-level lineage is real but narrow: BigQuery, and Dataproc since May 2026, while Dataflow, Composer, Vertex AI, Looker, and anything arriving through OpenLineage stay table-level. Documented limits worth checking against your workload include no column lineage for load jobs or routines, none upstream of external tables, top-level columns only with no nested STRUCT or JSON, and a silent fallback to table-level lineage once a job creates more than 1,500 column links.

AWS Glue Data Catalog and SageMaker Catalog

Two different layers, and the difference matters for a governance evaluation. The Glue Data Catalog is a metadata store holding databases, tables, and schemas for Athena, EMR, and Redshift Spectrum, and despite a vague marketing line in its own documentation it has no lineage user interface, no lineage API, and no column-level lineage. Governance features such as fine-grained access control come from Lake Formation layered on top. Column-level lineage lives in Amazon SageMaker Catalog, built on Amazon DataZone. Read its support matrix before relying on it: automation covers Glue crawlers and Redshift by default, while Glue console jobs and EMR need explicit configuration and Visual ETL in fine-grained mode is unsupported. DataZone was not renamed and still exists separately.