Testing in dbt - Tim Hiebenthal

Tim Hiebenthal is a Lead Analytics Engineer at Project A Ventures. In this presentation, he gives actionable insights on how to design tests for your dbt models.

1. Don’t publish new workflows without testing

2. Separate test environment from production environment

3. Assign severity levels to tests (e.g. error vs warning)

4. Go beyond dbt’s standard tests (unique, not null, relationships)

5. Monitor compute resources used by tests and reduce test cadence for less impactful tests

6. Have SOPs about who does what depending on test outcomes

7. Use tests to proactively catch business logic errors (mostly for manual data capturing processes, such as utm parameters)

8. Start with basic tests and iteratively add more complex ones as needed

9. The first custom tests to add should focus on testing data that is generated by humans.

10. Ensure that tests remain relevant and actionable

11. Join leading data experts and receive tips, strategies, and resources about how to make your CEO and Founders love you. Sign up below for FREE

Create more business impact with data & AI

Join 400+ leading data, analytics & AI experts and receive tips, strategies, and resources about building more impactful data teams and data infrastructure and how to make your CEO love you. 😍

✅ Free Newsletter

✅ Free Access to a vast knowledge base of actionable templates, best practice documents, and 25+ hours of video material

Error. Your form has not been submittedEmoji
This is what the server says:
There must be an @ at the beginning.
I will retry
Reply
I will never spam or sell your info. Promise.