Naming conventions at scale in dbt

Introduction

Good naming helps everyone understand and find data easily. It keeps projects clear and organized as they grow.

When working on a big data project with many tables and models.

When multiple people or teams collaborate on the same dbt project.

When you want to avoid confusion between similar data sources or tables.

When you need to automate or document your data pipeline clearly.

When maintaining data models over time and want to keep them consistent.

Syntax

dbt

-- Example naming pattern for dbt models
{{
  config(
    materialized='table',
    alias='stg_customers'
  )
}}

-- File name: stg_customers.sql

-- Table name: stg_customers

Use prefixes like stg_ for staging tables, int_ for intermediate, and fct_ for fact tables.

Keep names lowercase and use underscores to separate words for readability.

Examples

This creates a staging table named stg_customers to hold raw customer data.

dbt

-- Staging table for raw customer data
{{ config(materialized='table', alias='stg_customers') }}

select * from raw.customers

This creates a view named int_customers_clean with only active customers.

dbt

-- Intermediate table for cleaned customer data
{{ config(materialized='view', alias='int_customers_clean') }}

select * from {{ ref('stg_customers') }} where active = true

This creates a fact table fct_customer_orders summarizing orders per customer.

dbt

-- Fact table for customer orders
{{ config(materialized='table', alias='fct_customer_orders') }}

select customer_id, count(*) as order_count from {{ ref('int_customers_clean') }} join raw.orders using (customer_id) group by customer_id

Sample Program

This example shows a simple dbt project with three models using naming conventions: staging (stg_), intermediate (int_), and fact (fct_) tables. It helps keep the project organized and clear.

dbt

-- dbt model: stg_customers.sql
{{ config(materialized='table', alias='stg_customers') }}

select * from raw.customers

-- dbt model: int_customers_clean.sql
{{ config(materialized='view', alias='int_customers_clean') }}

select * from {{ ref('stg_customers') }} where active = true

-- dbt model: fct_customer_orders.sql
{{ config(materialized='table', alias='fct_customer_orders') }}

select customer_id, count(*) as order_count
from {{ ref('int_customers_clean') }}
join raw.orders using (customer_id)
group by customer_id

Important Notes

Consistent naming helps when searching or filtering models in dbt Cloud or IDEs.

Use clear prefixes to quickly know the purpose of each model.

Document your naming rules in your project README for team clarity.

Summary

Use simple, consistent prefixes like stg_, int_, and fct_ to organize models.

Keep names lowercase and separate words with underscores for readability.

Good naming makes large dbt projects easier to understand and maintain.

Practice

1. Why is it important to use consistent prefixes like stg_, int_, and fct_ in dbt projects?

easy

A. They make the SQL code run faster.

B. They help organize models by their role, making the project easier to understand.

C. They automatically create indexes in the database.

D. They allow dbt to skip testing models.

Solution

1. Step 1: Understand the purpose of prefixes in dbt

   Prefixes like stg_ (staging), int_ (intermediate), and fct_ (fact) indicate the role of each model.

2. Step 2: Recognize how prefixes help in large projects

   Using these prefixes organizes models logically, making it easier to find and maintain them.

3. Final Answer:

   They help organize models by their role, making the project easier to understand. -> Option B

4. Quick Check:

   Consistent prefixes = better organization [OK]

Hint: Remember: prefixes show model roles clearly [OK]

Common Mistakes:

• Thinking prefixes affect performance
• Assuming prefixes create database indexes
• Believing prefixes skip tests

2. Which of the following is the correct way to name a staging model for customer data in dbt?

easy

A. stgCustomerData

B. StgCustomerData

C. STG_CUSTOMER_DATA

D. stg_customer_data

Solution

1. Step 1: Recall dbt naming style guidelines

   dbt recommends lowercase letters and underscores to separate words for readability.

2. Step 2: Evaluate each option

   stg_customer_data uses lowercase and underscores, matching the recommended style.

3. Final Answer:

   stg_customer_data -> Option D

4. Quick Check:

   Lowercase + underscores = correct naming [OK]

Hint: Use lowercase and underscores for dbt model names [OK]

Common Mistakes:

• Using uppercase letters
• Using camelCase instead of underscores
• Starting with uppercase letters

3. Given the following dbt model names, which one is a fact table according to naming conventions?

stg_orders
int_customer_info
fct_sales_summary
stg_products

medium

A. stg_orders

B. int_customer_info

C. fct_sales_summary

D. stg_products

Solution

1. Step 1: Identify prefixes and their meanings

   stg_ means staging, int_ means intermediate, fct_ means fact table.

2. Step 2: Match the prefix to the fact table

   Only fct_sales_summary starts with fct_, indicating a fact table.

3. Final Answer:

   fct_sales_summary -> Option C

4. Quick Check:

   Prefix fct_ = fact table [OK]

Hint: Fact tables start with fct_ prefix [OK]

Common Mistakes:

• Confusing staging and fact prefixes
• Ignoring prefix meanings
• Choosing intermediate models as fact tables

4. You see a dbt model named fct_customer_orders but it contains only raw data loading logic without aggregations. What is the likely issue?

medium

A. The model name prefix fct_ is incorrect for raw data; it should be stg_.

B. The model should be renamed to int_customer_orders because it is intermediate.

C. The model name is fine; raw data can be in fact tables.

D. The model should be deleted because it is redundant.

Solution

1. Step 1: Understand prefix meanings

   fct_ is for fact tables, which usually contain aggregated or transactional data, not raw loads.

2. Step 2: Match model content to prefix

   Since the model only loads raw data, stg_ (staging) prefix is more appropriate.

3. Final Answer:

   The model name prefix fct_ is incorrect for raw data; it should be stg_. -> Option A

4. Quick Check:

   Raw data = stg_ prefix [OK]

Hint: Match prefix to model content type [OK]

Common Mistakes:

• Assuming any prefix works for any model
• Thinking raw data belongs in fact tables
• Renaming without checking model logic

5. In a large dbt project, you want to create a naming convention that helps new team members quickly identify model types and their dependencies. Which approach best supports this goal?

hard

A. Use consistent prefixes like stg_, int_, fct_, keep names lowercase with underscores, and document the naming rules.

B. Use random prefixes for variety, mix uppercase and lowercase, and avoid documentation to encourage exploration.

C. Name all models with the project name prefix only, ignoring model roles and dependencies.

D. Use only numbers as prefixes to indicate model order, without descriptive names.

Solution

1. Step 1: Identify best practices for naming at scale

   Consistent prefixes and lowercase with underscores improve clarity and readability.

2. Step 2: Recognize importance of documentation

   Documenting naming rules helps new team members understand conventions and dependencies quickly.

3. Final Answer:

   Use consistent prefixes like stg_, int_, fct_, keep names lowercase with underscores, and document the naming rules. -> Option A

4. Quick Check:

   Consistency + documentation = easier onboarding [OK]

Hint: Combine clear prefixes with documentation [OK]

Common Mistakes:

• Ignoring documentation
• Using inconsistent naming styles
• Choosing unclear or numeric-only prefixes