Why You Should Connect GA4 to BigQuery (Even on the Free Plan) | Ron Kopelman

The BigQuery Export: GA4’s Best-Kept Secret

When Google launched GA4, a feature flew under the radar for many marketing teams: the free BigQuery export. In Universal Analytics, this option was reserved for GA360, the paid version at 150,000 dollars per year. With GA4, everyone has access. And yet, a minority of businesses use it.

That is a mistake. BigQuery is probably GA4’s most powerful feature, and it costs virtually nothing for most websites.

What BigQuery Gives You

Raw data, event by event. In the GA4 interface, you see aggregates — totals, averages, percentages. In BigQuery, you access every individual event with all its parameters. You can reconstruct a user’s exact journey, analyze event sequences, or create custom metrics impossible to obtain in the interface.

No sampling. GA4 samples data when volumes exceed certain thresholds. BigQuery contains all of your data. For ad-hoc analyses or monthly reports, this guarantees total accuracy.

Analyses impossible any other way. What is the median time between first visit and purchase? What are the most common navigation paths before a conversion? Which users viewed the pricing page but never converted? These questions require SQL queries that the GA4 interface cannot support.

A Concrete SQL Example

To calculate the number of days between the first visit and the first conversion, the query looks like this:

SELECT
  user_pseudo_id,
  DATE_DIFF(
    MIN(IF(event_name = 'purchase', event_date, NULL)),
    MIN(event_date),
    DAY
  ) AS days_to_purchase
FROM `your-project.analytics_XXXXXXX.events_*`
WHERE _TABLE_SUFFIX BETWEEN '20260201' AND '20260228'
GROUP BY user_pseudo_id
HAVING days_to_purchase IS NOT NULL

This type of analysis is inaccessible in the GA4 interface but trivial in SQL once the data is in BigQuery.

What It Actually Costs

The GA4 to BigQuery export is free. BigQuery storage is free up to 10 GB per month. Queries are free up to 1 TB of data processed per month. For a site with normal traffic (a few hundred thousand sessions per month), you will likely stay within the free limits or pay a few euros per month.

The real cost is not financial — it is human. You need someone who can write SQL and understand the GA4 data schema in BigQuery. This schema is particular: data is nested (nested/repeated), which confuses analysts accustomed to classic relational tables.

How to Set Up the Export

Configuration takes ten minutes in the GA4 admin interface. Go to Admin, then the “Product Links” section, click “BigQuery Links,” and follow the wizard. You need a Google Cloud Platform project (free to create) and to enable the BigQuery API.

The export creates one table per day (format events_YYYYMMDD) plus an intraday table updated continuously. Historical data is not exported retroactively: the export starts the day you activate it. That is why you should enable it as soon as possible, even if you do not plan to use BigQuery immediately.

Where to Start

If you have not yet activated the export, do it today. The data accumulated over months will have considerable value the day you want to run a deep analysis. And if you need help structuring your first queries or connecting BigQuery to Looker Studio, a one-time support engagement can save you several weeks of learning.