Collibra is the grown-up in the data catalog room. Founded in 2008 — two years before Instagram, four years before Alation, and eleven years before Atlan — Collibra predates the entire modern data catalog category. It was spun out of the STARLab research group at Vrije Universiteit Brussel by four Belgian academics, led by Felix Van de Maele, who had been working on semantic web and ontology problems for enterprise data.

This origin matters. Collibra was not built for analysts trying to find a dbt model; it was built for data governance committees at European banks trying to comply with Basel II and, later, GDPR. The product grew up shaped by that customer base: heavy on policy workflows, rich in stewardship roles, deeply extensible via custom metamodels, and until very recently, extremely willing to take six months to deploy.

For most of the 2010s, Collibra was the safe enterprise answer. By 2022, it was a $5.25 billion private company with hundreds of Fortune 1000 customers, a European pedigree that made it the default pick for regulated industries on both sides of the Atlantic, and a reputation as the Oracle of metadata: expensive, powerful, and almost impossible to displace once deployed.

What Collibra Actually Does

Collibra's product is best understood as a governance platform that includes a catalog, rather than a catalog that happens to do governance. The distinction shapes almost every design decision.

Data Intelligence Cloud. Collibra markets its full suite as the "Data Intelligence Cloud," a bundle that includes:

• Collibra Catalog — the inventory, search, and lineage layer, most directly comparable to Atlan or Alation.
• Collibra Data Governance — the policy, workflow, and stewardship engine: assign owners, define policies, route approvals, run certifications, track issues. This is the heart of the product and the reason regulated customers buy it.
• Collibra Data Quality & Observability — born from Collibra's 2021 acquisition of OwlDQ, it adds rule-based and ML-driven quality checks on top of the catalog.
• Collibra Protect — policy-based access control that pushes masking and row-level rules down into connected warehouses (directly competing with Immuta and Privacera; see the Governance & Security section).
• Collibra Lineage — an increasingly serious lineage engine, including column-level lineage added in recent releases.

The Operating Model for Data Governance. Collibra's most distinctive intellectual contribution is what it calls the Operating Model: a configurable metamodel that lets an organization define its own governance vocabulary (domains, business assets, technical assets, roles, workflows, responsibilities) and have the product enforce it. In practice, this is why Collibra deployments take months: the first phase is a consulting engagement to design the operating model. In exchange, the organization gets a catalog that reflects how it actually works, not how the vendor thinks governance should work. For a global bank, this flexibility is worth the pain. For a 200-person startup, it is a nightmare.

Workflow engine. Collibra includes a full BPMN-style workflow engine for approvals, certifications, requests, and escalations. Every catalog has "workflows" as a checkbox; Collibra's is the only one a compliance officer actually respects.

Business glossary. The glossary is one of the original Collibra features and remains best-in-class. Terms can have formal definitions, approval states, multiple authorized users, rich relationships to other terms, and links to technical assets. For a regulated bank with 15,000 defined business terms, this is not optional.

The Opinionated Take

Collibra is fading from new cloud-native deals but remains deeply entrenched in finance, insurance, pharma, and government — and will be for a long time. This is the honest two-sentence summary, and both halves matter.

The fade is real. Walk into any Series-B to Series-D cloud-native company shopping for a catalog in 2025 and the shortlist is usually Atlan, DataHub, and maybe Select Star. Collibra rarely makes it past the first call; its sales motion is tuned for million-dollar enterprise deals and multi-month POCs, not for a Head of Data who wants a working product by next quarter. The UI still feels like enterprise software from 2015. Column-level lineage exists but is widely regarded as less polished than Atlan's or DataHub's. The dbt and Snowflake integrations work but are not the love letters that Atlan's are.

The entrenchment is also real, and it is the reason Collibra is still a multi-billion-dollar company. Regulated industries buy governance, not catalogs — and on pure governance weight, nobody in the modern wave has caught up. Collibra's policy workflows, operating model, glossary, and audit trail are genuinely deeper than anything Atlan ships. When a tier-1 bank gets a letter from the ECB asking how it tracks data lineage for regulatory reporting, the existing Collibra deployment is the answer, and no CDO is going to rip it out to save a few hundred thousand dollars a year.

The strategic question for Collibra is whether it can cross the chasm in the other direction: win new cloud-native deals without losing the enterprise feature depth that makes it defensible. The company has been investing in cloud deployment, AI governance (a hot topic as enterprises try to register and govern their LLM use), and faster time-to-value. Results so far are mixed.

The honest prediction. Collibra will remain the default catalog at global banks, pharma companies, and regulated government buyers for the foreseeable future. It will continue to lose the new-stack darling deals to Atlan. Over time it may become the "IBM of catalogs" — huge, profitable, boring, and somehow still in every RFP — or it may find a second act in AI governance, a category where its operating model and workflow engine are surprisingly well-suited to the emerging problem of tracking model inputs, datasets, and decisions.