This blog post is my personal experience through achieving the dbt Analytics Engineering Certification, with some tips here and there that can help you smash it.
What the heck is even the dbt Analytics Engineering Certification?
If you are on this page, you probably already know what this certification is. For those who don’t, the dbt Analytics Engineering Certification Exam assesses your proficiency in using the dbt platform, including both Cloud and Core, through a series of multiple-choice questions. The objective is to make you certified in building the T part (transformation) of ETL pipelines through different Data Warehouses, such as Snowflake, Databricks, or Amazon Redshift.
Topics and Suggestions
You can view exam topics here and you can find the Official Study Guide here. I recommend starting with the courses offered on the dbt portal. They are well-designed and provide hands-on experience with a fictional project, which can help you gain proficiency in using the dbt platform.
To develop proficiency in both dbt Cloud and Core, I suggest developing a new project using also dbt Core, as the exam focuses on both platforms. To understand new functions and ways of structuring the code, which are essential for the Certification, you should develop proficiency in them.
Be sure to experiment with concepts such as incremental models, seeds, macros and hooks. Even simple pieces of code have such a variety of settings and meaning behind them.
The Certification is technology agnostic, so you can choose whatever Data Warehouse platform you prefer: Snowflake, Redshift, Databricks, etc. A trial account is enough to develop a data pipeline and to test your knowledge.
Exam Structure
The exam consists of 65 multiple choice questions and the passing score is 65%. You can go back and forth between them. They can have one or multiple answers and there is a small percentage of questions, which I call the anxiety-provoking ones.
While the normal multiple choice questions aren’t much of a challenge, the other ones are. They basically consist of a flow of True/False questions related to a main one. You can’t go back and forth between them and once you provide a wrong answer the flow is interrupted.
Since you can have flows containing between two and six answers, even reaching the end with all correct answers can make you feel uncomfortable because of the sudden closing of the flow. You’ll be always asking yourself whether you successfully ended the flow or you provided a wrong answer.
You are provided with sufficient time to answer the questions on the dbt Analytics Engineering Certification Exam though, so you can take your time to answer them.
A real person will be proctoring you during the exam, who will be able to see your behavior thanks to a Zoom call in which you’ll have to share your entire screen. The cost is around $200 and the certification lasts 2 years.
Get Aways
In the end passing the exam is not that difficult but you need to prepare very thoroughly. They recommend 6+ months working with both dbt Core and Cloud but I was able to be certified in just one month, working on personal projects and studying the documentation, which is very detailed and well organized.
Unfortunately there are no valid mock exams to prepare on, but there is this website, which is free, where you are provided with a lot of questions on all the subjects of the exam (and even more). They are not related 1:1 to the actual exam ones, but they are a good way to revise your knowledge. There’s also a discussion section where you can talk with other to compare answers.
NOTE: by the time I’m writing this article, everything related to metrics and the semantic layer, is not part of the exam.
