dbt Cert Prep - Experience Needed.md

What Experience Do You Need for the dbt Certification?

📝 Study asset for the dbt Analytics Engineering Certification Practice Tests course on Udemy — 195 exam-style questions across 3 timed tests, updated for dbt 1.11.

Official Prerequisites

dbt Labs recommends:

• SQL proficiency — Joins, aggregations, CTEs, window functions
• 6+ months of experience building with dbt Core or dbt Cloud

What "6+ Months" Actually Means

The official recommendation is the bare minimum. Here's what practitioners who passed (and failed) report:

The Ideal Candidate

• 1+ years of hands-on dbt experience in a production environment
• Has built and deployed a dbt project end-to-end (not just inherited one)
• Has experience with debugging production failures — not just writing models
• Has worked with at least 3-4 materializations (not just tables and views)
• Has implemented tests beyond the basics (custom generic tests, not just not_null/unique)
• Has collaborated on dbt via Git (branching, PRs, code review)

Skills That Surprised People on Exam Day

Many experienced engineers reported being caught off guard by these topics:

Topic	Why It's Tricky
Snapshots	Most people set them up once and never think about them again. The exam asks about timestamp vs check strategy nuances, invalidate_hard_deletes, and what happens when check_cols is combined with timestamp strategy.
Model governance	Reworked in v1.11 (April 2026). Contracts (enforced: true, every column needs data_type), YAML column-level constraints (primary_key/not_null/check), versions (deprecation_date, latest_version). access levels and groups are no longer in scope. Many teams haven't adopted constraints yet.
State selectors	state:modified, state:new, result:error — the Slim CI workflow. Essential for the exam, rarely used hands-on by junior engineers.
Python models	def model(dbt, session) syntax, returning DataFrames, when to use Python vs SQL. Most dbt users have never written a Python model.
dbt clone	Zero-copy clones for CI/development. A newer feature most people haven't used.
Compiled SQL debugging	The workflow of checking target/compiled/ and target/run/ to diagnose issues. People debug differently in practice.
YAML syntax from memory	The exam is closed-book. Can you write a source freshness config, an exposure definition, or a custom generic test from memory? Most people can't.
Exact command flags	--defer, --favor-state, --full-refresh, --select state:modified+ — exact syntax matters.

Experience Level vs. Expected Pass Rate

Based on community reports from the dbt Slack #dbt-certification channel:

Experience Level	Typical Score	Notes
< 6 months dbt	40-55%	Not ready. Need more hands-on time.
6-12 months dbt, no study	50-60%	Fails. Experience alone isn't enough.
6-12 months dbt + 4 weeks study	65-75%	Passes, but barely.
1-2 years dbt + 2 weeks study	70-80%	Comfortable pass.
2+ years dbt + thorough study	80-90%	Strong pass.
2+ years dbt, no study	55-70%	Hit or miss. Often fails on governance/state.

Key insight: The #1 predictor of passing is dedicated study time, not years of experience.

SQL Skills Required

You don't need to be a SQL wizard, but you must be comfortable with:

• SELECT, FROM, WHERE, GROUP BY, HAVING, ORDER BY
• All join types (INNER, LEFT, RIGHT, FULL OUTER, CROSS)
• CTEs (WITH ... AS)
• Window functions (ROW_NUMBER(), RANK(), LAG(), LEAD(), SUM() OVER)
• Aggregations and CASE WHEN
• Subqueries
• UNION vs UNION ALL

The exam won't test advanced SQL like recursive CTEs or complex pivots, but it will present SQL in the context of dbt models and expect you to read it fluently.

Git Skills Required

You need working knowledge of:

• git clone, git pull, git push, git fetch
• The difference between git pull and git fetch (this is a real exam question)
• Branching: git checkout -b, git merge
• Pull request workflow
• Branch protection and code review concepts
• How dbt's development-to-production workflow maps to Git branches

You do NOT need to know advanced Git (rebasing, cherry-picking, bisecting).

What You DON'T Need

• dbt Cloud knowledge — The exam is dbt Core focused. No dbt Cloud-specific questions.
• Specific warehouse expertise — Questions are warehouse-agnostic. You don't need to know Snowflake or BigQuery specifics.
• Orchestration tool knowledge — No Airflow, Dagster, or Prefect questions.
• BI tool knowledge — No Tableau, Looker, or Preset questions.
• Advanced Python — Python models are covered, but you just need to know the basic def model(dbt, session) pattern.

Self-Assessment Checklist

Before registering for the exam, can you answer YES to all of these?

• I can write a dbt model from scratch using ref() and source()
• I can explain the difference between all 4 materializations and when to use each
• I can configure an incremental model with unique_key and is_incremental()
• I can set up snapshot tracking using both timestamp and check strategies
• I can write a custom generic test as a macro
• I can debug a dbt error by reading compiled SQL in target/compiled/
• I can write a source freshness configuration from memory
• I can write an exposure YAML definition from memory
• I can explain model contracts (enforced: true + data_type requirement), YAML column-level constraints, and versions (deprecation_date)
• I can explain state:modified+ and how Slim CI works
• I can use dbt retry and explain what it reads to know what failed
• I can write a dbt_project.yml configuration from scratch
• I know the difference between var() and env_var()
• I know how generate_schema_name works by default and how to customize it

If you answered NO to 3+ items, spend more time studying those areas before booking the exam.

Updated to the 2026 dbt v1.11 Certification