Advanced dbt-GA4 (On-Demand)

Go beyond the basics of the dbt-GA4 package. Build the architectural foundation you need to lead data engineering projects and advance your career.

Why “Advanced” Matters in the Age of AI

If you ask an LLM how to customize a model in dbt-GA4, it will likely give you code that “works.” But “working code” is the bare minimum.

I’ve seen AI prompts suggest painful paths—suggesting edits to source seeds that break on every package update, or providing confident, “probabilistic” guesses that are absolutely wrong.

AI can write the SQL. But the real, career-sustaining value is in understanding the connections between the technology, the stakeholders, and the architecture.

To help you focus on the value portions of this work, you will get access to my personal Cursor AI templates at the end of the course. These are markdown AI scripts for building dimensional models and then building aggregate models on top to meet report requirements extracted from report mockups.

This course isn’t just a collection of tutorials; it’s a sandbox designed to teach you how to think like an Analytics Engineer.

The Experience Behind the Curriculum

The Advanced dbt-GA4 course is built on my experience as one of the lead authors of the dbt-GA4 package and dozens of real-world implementations. I’ve architected GA4 solutions for:

• High-growth B2B SaaS companies requiring complex attribution logic.
• Fast-paced, e-commerce brands focused on sales in different markets around the world.
• Global non-profits managing data across dozens of international properties.

I’ve put over 200 hours into this curriculum to move you past “running scripts” and into architecting solutions.

Is This Right For You?

This is an advanced, hands-on track. To follow the exercises and get the most out of the curriculum, you must have:

1. A BigQuery Project with a Billing Account: The dbt requires specific permissions that the free sandbox disables.
2. A Live GA4 Data Stream: You need real, event-level data to see how these models behave in production.

Why can’t I use the public dataset? The public ga4_obfuscated_sample_ecommerce dataset will not work for this course. The dbt-GA4 package uses a date lookback (e.g., “last x days”) to process data. Because the public data is static and several years old, the models will fail to find new records to process.

What about the setup? We do a light review of the basic setup covered in the dbt-GA4 Setup course.

• If you took the Setup course: You can continue using that same project—just skip the tasks you’ve already completed.
• If you are an experienced dbt user: You might be able to skip the Setup course, provided you are using the native GA4 BigQuery integration (configured directly in the GA4 interface). Third-party API exports offer pre-aggregated stats; we need the raw, event-level granularity of the native export.

Finally, this is not a SQL course. You don’t need to be an SQL expert. But you can’t be afraid of it either.

What You’ll Learn

By the end of this course, you will be able to:

• Architect solutions on the dbt-GA4 foundation: Learn to use dbt as it was intended. Configure package internals and override models with intention so your project remains scalable and upgradeable.
• Lead the professional implementation lifecycle: Move from gathering requirements to facilitating high-stakes IT kickoff calls. Translate abstract business requirements into concrete data models using a “plan-first” approach.
• Develop core analytics engineering skills: Build a stable architectural base across Git, dbt, and BigQuery. Gain the confidence to tackle technical challenges independently, solving problems that go beyond any specific tutorial.
• Maximize performance and cost-efficiency: Master partitioning, clustering, and cost-monitoring in BigQuery to ensure your data stack stays fast and your cloud bill stays low.

Log In to Purchase

How You’ll Learn: The Four-Part Framework

We don’t do “copy-paste” learning. Every lesson in this course is structured to ensure you don’t just see the concepts, but master them through application.

1. Review: We start every module by revisiting previous content, ensuring you understand the foundation we are about to build upon.
2. Lesson: A strategic mix of deep-dive video and text walkthroughs with some supporting resources and templates.
3. Practice: Hands-on, basic exercises where you immediately build on the lesson content within your own dbt project.
4. Knowledge Expansion: The “Real World” layer. You’ll apply what you’ve learned to wider business contexts, solving problems that haven’t been “pre-solved” for you.

We’ve structured the curriculum to take you step-by-step from a basic setup to a fully-custom, production-grade dbt project.

Each chapter provides one practical goal that you will work towards. While doing so, you will review previous lessons, learn important new concepts, and then practice and apply them.

Bonus: Data Engineering Cursor Pack

Advanced dbt-GA4 teaches a structured approach to building models to meet report requirements.

Do you know what does well when working in a well-structured environment?

Coding LLMs.

That’s why I’m including access to my Cursor Data Engineering templates. These are currently focused on Airbyte and dbt on BigQuery and they slowly get improved every time that I use them. But they are currently, and probably always will be, a work-in-progress so they aren’t for beginners.

This is on top of Jupyter Notebooks that I share during the course that do things like extracting all events and event parameters from BigQuery helping you know exactly what GA4 settings you need to configure when setting up dbt-GA4.