tDS Data Catalog

A spreadsheet that documents your whole data estate — tracked events, prioritized metrics, tables, and columns — in one place. Built on the public TheLook e-commerce dataset so you can see exactly how it's done.

Primary Outcome

A single, shared source of truth for what data you have, how it's defined, and which metrics matter most.

Solution

A four-tab catalog — events, metrics, tables, columns — that becomes the single source of truth for what data exists, how it's defined, and what matters most.

Deliverables

  • An Events sheet documenting every tracked event, its trigger, and its properties

  • A Metrics sheet with prioritized metrics (P1–P5), categories, dimensions, and definitions

  • A Tables sheet describing each warehouse table and its grain

  • A Columns sheet with column-level descriptions, data roles, and keys

  • A worked example built on the public TheLook e-commerce dataset

Strategic Context

Undocumented data is data nobody trusts. A catalog is the difference between an analyst answering a question in five minutes and spending a day reverse-engineering a table. It's also where you settle metric definitions — the arguments that quietly waste hours of every leadership meeting.

Technical Architecture

Undocumented data is data nobody trusts. A catalog is the difference between an analyst answering a question in five minutes and spending a day reverse-engineering a table. It's also where you settle metric definitions — the arguments that quietly waste hours of every leadership meeting.

Problem Statement

Tables go undocumented, metrics mean different things to different people, and analysts waste hours reverse-engineering data nobody can explain.

Links

What's Included

A multi-tab Google Sheet: Events, Metrics, Tables, and Columns, pre-filled with the TheLook e-commerce example, including BigQuery INFORMATION_SCHEMA exports you can regenerate for your own warehouse.

FAQs

Do I have to use BigQuery?

No. The Tables/Columns tabs are easiest to auto-fill from BigQuery, but the catalog structure works for any warehouse.

How is this different from a data dictionary?

It's broader — it adds tracked events and a prioritized metrics layer, not just column descriptions.

Tech Stack

Tool 1

Tool 4

Tool 4

Tool 3

Tool 2

Tool 4

Primary Outcome

A single, shared source of truth for what data you have, how it's defined, and which metrics matter most.

Problem Statement

Tables go undocumented, metrics mean different things to different people, and analysts waste hours reverse-engineering data nobody can explain.

Solution

A four-tab catalog — events, metrics, tables, columns — that becomes the single source of truth for what data exists, how it's defined, and what matters most.

Links

Deliverables

  • An Events sheet documenting every tracked event, its trigger, and its properties

  • A Metrics sheet with prioritized metrics (P1–P5), categories, dimensions, and definitions

  • A Tables sheet describing each warehouse table and its grain

  • A Columns sheet with column-level descriptions, data roles, and keys

  • A worked example built on the public TheLook e-commerce dataset

What's Included

A multi-tab Google Sheet: Events, Metrics, Tables, and Columns, pre-filled with the TheLook e-commerce example, including BigQuery INFORMATION_SCHEMA exports you can regenerate for your own warehouse.

Strategic Context

Undocumented data is data nobody trusts. A catalog is the difference between an analyst answering a question in five minutes and spending a day reverse-engineering a table. It's also where you settle metric definitions — the arguments that quietly waste hours of every leadership meeting.

FAQs

Do I have to use BigQuery?

No. The Tables/Columns tabs are easiest to auto-fill from BigQuery, but the catalog structure works for any warehouse.

How is this different from a data dictionary?

It's broader — it adds tracked events and a prioritized metrics layer, not just column descriptions.

Technical Architecture

Undocumented data is data nobody trusts. A catalog is the difference between an analyst answering a question in five minutes and spending a day reverse-engineering a table. It's also where you settle metric definitions — the arguments that quietly waste hours of every leadership meeting.

Tech Stack

Tool 1

Tool 4

Tool 4

Tool 3

Tool 2

Tool 4