Skip to main content
Welcome to the Data Dictionary. Whether you’re looking up a metric, writing SQL, or understanding table structure, this is your hub for all things data at SourceMedium.

​
I want to…

Look up a metric

Find definitions for AOV, ROAS, LTV, CAC, and 180+ other metrics

Browse table schemas

Browse table schemas with column descriptions and data types

Write SQL queries

Copy proven SQL templates for common analyses

Access BigQuery

Learn how to connect and run your first queries

Understand data structure

Learn about facts, dimensions, OBTs, and SourceMedium’s modeling approach

Look up a term

Get definitions for channels, valid orders, and common SourceMedium concepts

​
Quick Reference

For fast lookups, use Metrics, Dimensions, or the Glossary.

​
Table Documentation

sm_transformed_v2

Recommended for most use cases — cleaned, enriched analytics tables (facts, dimensions, OBTs, reports)

Metadata Tables

Semantic layer catalog and automated data dictionary for programmatic lookups

Experimental Tables

Beta features including multi-touch attribution models and advanced attribution tables
New to SourceMedium’s data model? Start with Modeling Philosophy to understand how we structure data (facts, dimensions, OBTs, reports).

​
Working with Data

BigQuery Essentials

How to access your warehouse and run queries

SQL Query Library

Copy-paste SQL for cohort analysis, retention, product performance, and more

Common Analyses

Proven patterns for top products, ROAS, retention, LTV, and profitability

Connect BI Tools

Link BigQuery to Looker Studio, Tableau, or other visualization tools

​
Schema Standards

Naming Conventions

Learn the patterns behind table prefixes (fct_, dim_, obt_, rpt_), column naming, and metric calculations

​
Advanced Resources

Browse raw source schemas for integrated platforms (Shopify, Klaviyo, Meta Ads, Google Ads, etc.).When to use: Debugging integrations, understanding what SourceMedium transforms, or accessing platform-specific fields not yet in sm_transformed_v2.For most analyses: Use sm_transformed_v2 tables instead — they’re cleaned, enriched, and easier to work with.View Platform Raw Data →
Need custom tables or logic beyond SourceMedium’s out-of-the-box models? Learn how to extend the data warehouse with your own dbt models.Managed Data Warehouse Modeling Guide →