Learn
What is Data Catalogue?
An organised inventory of all the data in your organisation – what exists, where it lives, what it means, and who owns it – so people can find and understand data without asking around.
What it is
A data catalogue is like a library catalogue for your business data. Instead of books, it indexes tables, columns, dashboards, and reports. Instead of Dewey Decimal numbers, it provides descriptions, data types, owners, and usage statistics. Without a catalogue, finding the right data involves Slack messages, email chains, and hoping someone remembers which table the "real" revenue lives in. With a catalogue, an analyst can search for "monthly revenue," see every table that contains revenue data, read the definition of each one, and know which is the official source – all without asking anyone.
Why it matters for your business
As your data grows, findability becomes a bottleneck. Alation found that data professionals spend 30% of their time searching for and understanding data before they can start analysing it. That is nearly a third of your analytics team's capacity spent just looking for the right table. A $70M professional services firm we worked with had 450 tables in their warehouse. New analysts took three weeks to become productive because nobody had documented what each table contained. After building a catalogue with descriptions, owners, and lineage links, onboarding time dropped to four days.
How we approach it
We build catalogues that are practical, not encyclopaedic. Every gold-layer table gets a plain-English description, a defined owner, a freshness indicator, and links to the dashboards that consume it. We use dbt's built-in documentation features to auto-generate catalogue entries from model definitions, so the catalogue stays in sync with the actual data rather than becoming a stale wiki nobody trusts. For larger organisations, we integrate tools like Atlan or DataHub that add search, lineage visualisation, and usage tracking.
Related Services
Key Takeaways
- •A data catalogue lets your team find and understand data without relying on tribal knowledge or Slack messages.
- •Data professionals spend 30% of their time searching for data – a catalogue gives that time back to actual analysis.
- •The best catalogues are auto-generated from your data models, not manually maintained wikis that go stale.
- •Start with your most-used tables: descriptions, owners, and freshness indicators for your gold-layer models.