Preparing for the Updated dbt Analytics Engineering Certification Exam

Certifications are a great way to validate your expertise in specific tools or software — they make you or your team more credible in the eyes of leadership, peers, and potential employers.

When it comes to earning the dbt Analytics Engineering Certification, preparing for the exam can deepen your understanding of dbt and how you apply engineering principles to analytics infrastructure. dbt Labs made significant revisions to the exam in May 2024. If you’re a longtime dbt user, you may not be aware of dbt’s newer features, which the exam will definitely test you on.

After you become certified, you must renew your certification every two years. Many of our analytics engineers at BDC recently renewed their certifications, so we needed to update our internal resources to reflect the updated content on the exam. Here’s what we’ve learned has changed on the exam.

What’s New?

• Model governance — Understand dbt’s three ways of enforcing data governance: model contracts, model versions, and model access.

• State comparison — Know how dbt tracks state via its artifacts, how dbt retry relies on those artifacts, and how to use state-based node selectors.

• Python models — Explain how dbt can execute data transformations written in Python and key Python functions to implement these models.

• dbt clone — Describe how dbt can clone selected objects to target schemas (often, via zero-copy cloning if the data warehouse supports it), which is handy for reducing compute cost and maintaining a mature CI/CD workflow.

• grants — Use the built-in grants config instead of on-run-end hooks to manage user access to database objects at any project level.

• And some smaller items — exposures, source freshness, and custom generic tests. These features have been part of dbt for quite some time but have only recently made it into the exam!

What’s Been Dropped?

• Environments — Many aspects of environment management are no longer on the exam, including:

• Creating development vs. deployment environments
• Understanding the differences between production, development, and raw data
• Configuring dbt connections to environments in the data warehouse
• How the “generate_schema_name()” macro interacts with environments to produce custom schema names

• Model selection — The exam won’t ask you how to write dbt commands that execute a subset of models using selectors (but this is still one of dbt’s most important features and something you should be comfortable with).

What’s Been De-Emphasized But Still on the Exam?

• dbt docs — The much-beloved static-site documentation feature is no longer a focus given the recent release of dbt Explorer in dbt Cloud.

• Git workflows — The exam will ask far fewer questions about the “right” way to work with version control systems since we know now different teams using dbt will use various approaches for git branches, commits, and pull requests.

Remember, none of dbt’s paid features such as dbt Semantic Layer, dbt Explorer, dbt Cloud CLI, or dbt Mesh, are on the exam because the goal of dbt’s certification is to validate your expertise in developing essential data transformation workflows on dbt Core or in the dbt Cloud IDE.

Finally, don’t miss the Learning Path in dbt Labs’ official study guide — it’s a recommended curriculum for studying the various topics on the test and will set you up for success. Good luck with the exam!

Need more help prepping for the exam or using dbt in your data analytics stack? Reach out. Our data experts would be happy to assist your team.