Combining GA4 with other tools: BigQuery, data import, and Measurement Protocol

GA4 is a powerful tool alone, but the real business dataset is never alone. CRM stores customer loyalty metrics, product features in the warehouse system, and cash transactions at the offline point of sale. This article explains how to integrate GA4 with those systems — based on Google Analytics Academy's advanced course content.

Source: Google Analytics Academy course list

Why to merge the dataset

Each system you use creates its own dataset. GA4 aggregates the site and app data through tags and the Firebase SDK — but that's only part of the picture.

• CRM can have customer loyalty rates, lifetime value, and product benefits.
• If you're a web publisher, your content management system stores dimensions like author and article categories.
• If you're in e-commerce, product features like price, style, and size are in a separate system.

This is illustrated by an example of Gamze in the course. Gamze has a photo editing app and website, analysing their use with GA4. But information about “pro” subscribers extends beyond Analytics. He wants to connect that dataset to the GA4 dataset — and there are three ways to do that.

Path 1: Change the Event and create a new event

The simplest way is to use the GA4 interface — without changing the site or app code.

Modify event (change existing event)

• Re-name the event to use the same event name between property and data streams.
• Fixing spelling errors caused by measurement error.

Create event

When site visitors view any page page_view the event is running, but you may want to track views of a specific page as a separate event. In this case, existing page_view from the event, you create a new event and run it only for that page. If the page is important enough, you can also mark the new event as a key event.

Important warning: when a new event is created and an action meets its criteria, that action may also meet the criteria of the original event. For example, for the homepage page_view_homepage you've created a new event called — when the user visits the homepage, both page_view_homepagebut also original page_view it works, and both are counted. If you ignore this in reports, the numbers will show up.

The real definition of this function is the available dataset expand it's all about creating, editing, and synthesizing key events before Analytics processing. The following are two ways to expand the dataset.

Path 2: Measurement Protocol

Measurement Protocol is a standard set of rules that enables events to be collected and sent directly to Google Analytics servers.

Google automatically sends events after the tag and Firebase SDK are applied. The Measurement Protocol is different — it requires you to program events manually. Developers set up an HTTP request and send events directly to GA4 servers.

This is necessary in cases where:

• Connecting online behaviors to offline behaviors
• Measuring both client-side and server-side interactions
• Sending events outside of the site and app — offline conversions, order refunds, in-store purchases
• Sending data from internet connected devices, such as a Kiosk or POS (point of sale) system

Technical mechanism: Measurement Protocol dataset with existing Analytics dataset user_id and event_id as long as the connection is extended through the switches. That is, events that already exist in GA4 and events brought by the Measurement Protocol are tied together.

Path 3: Data import

Data import allows you to load and merge or overwrite your external data with an existing GA4 dataset. The offline data you upload merges with the event dataset that GA4 collects, resulting in a wealth of your reports, comparisons, and audiences.

The technical side is simple: you upload your external device to the GA4 property in the form of CSV files. You can export these CSVs directly from an offline business tool like a CRM or CMS. For small amounts of data, the file can also be created manually in a text editor or in a table.

Among the types of data that can be imported are cost data (advertising costs), item data (product data), user data, and offline event datasets. The most common scenario is importing the cost data of advertising platforms (e.g. social networks) outside of Google and seeing the ROI of all channels in GA4 in one place.

Exporting GA4 dataset to BigQuery

The GA4 interface is strong, but has limits: ready reports show an aggregated dataset, can be sampling, and can't write SQL. BigQuery Export goes beyond just that.

What is BigQuery?

BigQuery is a serverless, affordable multi-volume data warehouse. It has the capabilities of machine learning, geographic data analysis, and business intelligence. It allows you to run SQL queries without managing infrastructure, and thanks to a scalable analysis engine, you're able to query large data sets.

GA4 → BigQuery flow

From GA4 properties raw event dataset You're exporting to BigQuery, which is a data that's not aggregated, with a separate line for each event — a level of detail that's not accessible on the GA4 interface. Then you can merge this dataset with external data or transfer it to external storage.

There are two options for export frequency and it is possible to select both at once:

• Daily — all data is exported once a day.
• Streaming — data is exported continuously throughout the day. Note: some data cannot be accessed immediately; for example, the user attribution dataset may take longer than 24 hours to process fully.

You also select the geographic area where the data will be stored — this is set during the BigQuery Export configuration in GA4. This is especially important for markets with data localization requirements.

Important update: BigQuery Export is now available for free properties

Previously, data exports from Google Analytics to BigQuery were only available to corporate Analytics 360 properties through Universal Analytics. BigQuery Export in association with GA4 both standard and 360 properties opened for us. This has opened a serious door in analytics for small and medium-sized businesses — that owning raw data no longer requires a corporate budget.

Test-ready data sets

If your own data is still low, you can start with Google's sample data sets. There are two sets: the e-commerce site web dataset and the game app application dataset. With these, you can test ready-made sample requests — both basic and advanced.

Control over how the data will be used: privacy settings

Data protection standards have been rising as consumer expectations have increased and legislation has changed. GA4 has several tools to meet these requirements.

Data collection control

• Turning off data collection — you can partially or completely stop data collection in your digital properties.
• IP Addresses — GA4 does not individually record or store IP addresses. That means you don't need to change additional settings in this direction.
• Data retention (retention period) — set how long the user and event level data will be stored before automatic deletion. Note: this setting does not affect reports based on aggregated datasets.
• Consent mode — applied Google tags are dynamically matched by users' permission status. When the user grants permission, the tags use the measuring instruments only for the specified purposes.

Removing data from GA4

If you have mistakenly collected personal identification information (PII) or need to delete a dataset for any reason, there are several options:

1. Data deletion request — You request data deletion from Analytics servers. Before deletion begins 7 days the waiting period is granted; in the meantime, all users and administrators with editing permission are notified and can cancel the request.
2. Deleting single user data — The user in Explore is done with the exploration technique. Data related to that user 72 hours removed from the report within, and removed from servers in the next deletion process.
3. Removing a property — after all reports of the property and its contents have been moved to the trash 35 days permanently deleted after deletion. Historical data cannot be recovered after deletion.
4. User Deletion API — to delete the data in bulk. Used when multiple users need to delete their data at once.

Caution: reports based on aggregated datasets (e.g. user counts in Audience overview) are not affected by these deletions.

Ad personalization

When you enable ad personalization in Analytics (for example, by enabling Google signals), you can use your GA4 audience to personalize your digital ads. This setting can be controlled in two levels: at property level (turn off for entire property) and country-level (for specific countries).

Analytics 360: corporate opportunities

Analytics 360 is a paid, corporate version of Google Analytics. GA4 builds on properties and meets the demands of big advertisers, agencies: more personalization, increased scalability, and corporate support.

The most important difference is the two additional property types in the account structure:

• Subproperty — Retrieves data from another property (source property). Typically the source is a subplot of the property's data.
• Roll-up property — merges two or more source property datasets. Can contain data of ordinary properties and subproperty, but not data of another roll-up property.

When you need a subproperty

Example of the course: Anil is the head of a large insurance company and the company offers three types of polices: home, car, life insurance. These directions should be analyzed separately, but they should also be looked at together as they complement each other. Anil sends the entire product dataset to the same source property, then creates a subproperty for each business direction.

The second widespread use case is user management: for example, if a team in North America is required to see only the data of their region, while a dataset in South America is required to be linked to them, then subproperty is the ideal solution.

When you need a roll-up property

The second example in the course: Hilal is the owner of a retail company and the company has three independent brands — automotive products, home goods, home electronics. Each brand operates independently and needs to analyze its data separately, but the parent company wants to see the total number of users and revenue in one place. Solution: take three brand properties as sources and create one roll-up property.

Combining GA4 with other Google tools

Gamze uses a variety of tools to manage its digital ads, grow its app, and measure its site's search traffic. Let's look at the advantages of connecting them with GA4.

Google Ads

Benefits to the Google Ads side: conversions from GA4 are used for reporting and Smart Bidding (ROI increase); audiences are available for reporting and remarketing when Google signals are enabled.

Benefits on GA4 side: promotion metrics (clicks, costs) and dimensions (campaign name, keyword) are generated automatically; Google Ads campaign dataset is included in attribution reports.

Display & Video 360

Display and video is a platform that manages campaigns, offers, promotional materials, and audiences. After the connection, conversions and audience are available on the DV360 side; on the GA4 side, DV360 ads appear in cross-channel reports, explorations, and attribution, and the promotion and expense data is transmitted to GA4.

Search Ads 360

Google Ads is for managing search campaigns on Bing, Yahoo, Baidu, and other platforms. When connected to GA4, the multi-platform search dataset is merged into a single report — this is critical for advertisers, not just those with Google Ads.

Besides these, GA4 can be combined with Search Console (organic search dataset), Firebase (mobile app), Google Play, and BigQuery.

Frequently asked questions (FAQs)

Is BigQuery Export for paid properties?

No. Together with GA4, BigQuery Export is available for both standard and Analytics 360 properties. The data limits for property types differ.

Can I select the location of the data sent to BigQuery?

Yes. You set the geographic location where the data will be stored during the BigQuery Export configuration in GA4.

Can I cancel a data deletion request?

Yes. A 7-day waiting period is granted before removal begins, and users with editing permission may cancel the request during this time.

What is the difference between Measurement Protocol and data import?

The Measurement Protocol sends events in real time, directly to servers with an HTTP request, and requires developer work. Data import, however, loads external data in bulk with CSV files and requires less technical knowledge.

What is the difference between a subproperty and a roll-up property?

The subproperty takes the dataset from a source property and shrinks it (creates a subroutine cluster). The roll-up property combines the data of several source properties. Both are Analytics 360 capabilities.

Results

The true strength of GA4 is revealed when combined with other systems. For simple needs, it's enough to create an event over an interface; for offline and server dataset, the Measurement Protocol is needed; for CRM and cost dataset, data import runs; for raw data and SQL analysis, BigQuery Export provides the key.

Privacy settings — data retention, consent mode, deletion requests — ensure the legal security of this entire structure. If you follow these steps in the right order, GA4 goes from being just a reporting tool to becoming the data center of your business.

👉 Key event, custom report, and audience in GA4

Source: Analytics Academy — “Use GA with other tools and data sources” course