Introduction

By Timo Dechau · Last updated March 25, 2026

So you’ve enabled the GA4 BigQuery export. Nice. You’re already ahead of most teams — you’ve got raw event data sitting in BigQuery, ready to be queried.

Now here’s the thing: the gap between “export is enabled” and “we have a reliable data model we actually trust” is where teams lose weeks. Sometimes months. I’ve seen it happen over and over again across client engagements.

The GA4 BigQuery export is genuinely powerful. But it’s also full of traps. Some are obvious once you hit them — like that first terrifying BigQuery bill. Others are subtle — the kind of thing where your numbers look plausible but are quietly wrong, and you don’t find out until someone asks a question you can’t answer.

This guide covers 14 pitfalls I’ve seen repeatedly. Not theoretical risks — actual things that actual teams run into. Each one follows the same structure:

• The Trap — what people do wrong and why it seems reasonable
• Why It Happens — what about GA4 or BigQuery makes this non-obvious
• The Fix — a concrete solution, usually with SQL you can run
• How to Check If You Have This Problem — a diagnostic query or quick check you can run against your own data right now

That last part is the important one. I didn’t want this to be something you read, nod along to, and forget. Every pitfall gives you something to actually do — run a query, check a setting, have a conversation with your team. That’s what makes it useful.

The pitfalls are organized in the order you’ll likely encounter them: infrastructure and cost surprises first, then schema and export fundamentals, session and identity challenges, a couple of mindset shifts that save enormous amounts of wasted effort, and finally the data modeling principles that tie it all together.

At the end, you’ll find the Attribution Readiness Scorecard — five questions to honestly assess where your organization stands on the road from “raw GA4 data in BigQuery” to “attribution we actually trust.” It’s a good conversation starter with your team.

Let’s get into it.

Note

A note on currency: GA4’s BigQuery export schema evolves. Everything in this guide is current as of March 2026. If you’re reading this later, the principles hold — but double-check the specifics against Google’s documentation.

Tip

A note on export limits: For standard (non-360) GA4 properties, the streaming export is limited to 1 million events per day. The daily export limit has been effectively removed for most users, though Google’s “best effort” policies still apply — very high-volume sites may encounter overflow tables. If you’re on a 360 property, these limits are significantly higher. Either way, check your event volume against these thresholds early, before you build pipelines that assume every event arrives.