Google Analytics 4 Properties: Frequently Asked Questions

March 03, 2021
Lead Analytics Consultant

The Bounteous team held an open Q&A session where anybody could ask a burning question about Google Analytics 4 (GA4) Properties (view the recording here). Organizations and analysts have a lot of questions about the latest version of Google Analytics, so we rounded up the best ones to share.

Before we get started, note that these questions came from people who are already familiar to various degrees with both Universal Analytics (UA) and GA4 Properties. If you're just getting started with what Google has described as the "future of Google Analytics," read this great introductory post from our Director of Google Analytics, Samantha Barnes before you come back to this article.

With that, let's get started.

What's the Limitation Number for GA4 Custom Events?

First, let's provide a bit of context.

In Universal Analytics properties, data is organized into Users > Sessions > Hits, and there are several hit types, such as pageviews and events.

In the GA4 data model, sessions are removed from the hierarchy and every hit is an event.

Google Analytics 4 Data Model

 

Counting sessions? Send a session_start event. Counting pageviews? Send a page_view event. Counting purchases? Send a purchase event. And so on.

Some events are automatically sent to Google Analytics. Other events are recommended by Google and need to be configured. You can also create custom events that you invent.

This question asks specifically about the number of custom events that can be sent.

In theory, you can collect up to 500 distinctly named events per client ID and 500 per app instance, regardless of whether they are recommended events or custom events. Automatically collected events and those that are collected via enhanced measurement do not count toward this limit.

In practice, we haven't seen this limit enforced yet; however, the non-enforcement of a rule is no excuse to not follow it. Future-proof your data by staying within the data collection limits.

What's a Client ID? What's an App Instance?

These questions weren't actually asked during our Q&A session, but we need to know what these terms mean in order for the previous answer to make sense.

Client ID and app instance might be new terms to you. In web data streams, when a user goes to your site, Google Analytics sets a cookie to recognize your particular browser and device and stores a random string of numbers to serve as an identifier. This is how Google Analytics recognizes that user as the same user as they navigate from page to page.

If the user deletes cookies and returns to your site, a brand-new cookie with a brand-new client ID will be saved for that device and browser combination and a new user will be counted (more on the basics of the Google Analytics cookie).

In app data streams, when your app is downloaded, it generates a similarly random identifier. When somebody uninstalls your app and reinstalls it, a brand-new app_instance_id is generated and they appear as a new user.

How Do You Move Events Tracking to GA4 Seamlessly and Add New Parameters?

Let's break this question down into a couple of pieces.

We're going to assume that you're already collecting events in Universal Analytics.

The first thing you need to do is map your existing events to the GA4 data model. We offer some event naming considerations in this article as you design your event schema.

Second, you need to tag your app or website with the new events. The process is different depending on whether you're using Google Tag Manager or the Global Site Tag (gtag.js) for your website. For your apps, you must install Firebase and log events that way.

Google has step-by-step guides to walk you through the tagging process for your website, and some general instructions for your app. The Firebase documentation offers more detail on logging events for both Android and iOS data streams.

What's the Best Practice When Building Out Custom Events or Custom Parameters on Recommended Events?

We have a blog about this very topic, Event Naming Considerations for Google Analytics 4 Properties!

When Will Dimensions Like Last-click and Last-engagement Source, Medium, and Campaign Be Available to Be Paired With Event Parameters?

We actually already have event-scoped campaign parameters for conversion events. If you need source, medium, or campaign information for a specific event, mark it as a conversion.

The event's campaign parameters will inherit the session_start's campaign parameters, not the user's acquisition parameters. Here's an example where we're looking at a single user (client_id is captured as a user property in this case, so we can add a comparison to the report and easily focus on this user) and a conversion event called file_download:

report shown from Engagement > Events, with Session source/medium applied as a secondary dimension

 

This report comes from the Engagement > Events, with Session source/medium applied as a secondary dimension. Notice right below it, the session_start event with the same session source/medium.

See the same event in the Conversions > file_download report:

Conversions > file_download report

 

You might be wondering whether the Event source/medium inherits the User source/medium rather than the Session source/medium. It does not:

report showing that the Event source/medium doesn't inherit the User source/medium

 

If the Event source/medium were to inherit the User source/medium, then the Conversions > file_download report would show its source/medium as google / cpc rather than the bounteous.com / referral as demonstrated in the previous screenshot.

How Does Attribution Change with GA4 from UA?

UA attributes goal completions to the last non-direct click in standard reports and features a Multi-Channel Funnel Model Comparison Tool to see how attribution across channel, source, medium, and campaign would change based on your attribution model.

By contrast, GA4 offers a Comparisons feature where you can change the attribution type reflected in the report. There are four attribution types:

image showing the Google Analytics 4  - Four Attribution Types
Cross-channel last click

This is the default selection. It gives clicks from all sources and attributes the conversion to the last non-direct (source) click.

Cross-channel last engagement

Similar to the above, but this type will also give credit to an ad impression if there is no recent click. AKA "view-through conversion."

Google Ads preferred last click

This type gives credit to the latest Google Ads click regardless of whether there was a more recent engagement via another source.

Google Ads preferred last engagement

This type gives credit to the latest Google Ads impression regardless of whether there was a more recent engagement via another source.

Will "Annotations" Be Added to GA4?

The Google Analytics blog is the best place to get visibility into the GA4 feature roadmap.

Does GA4 Support a Referral Exclusion List Like Universal Analytics?

Not right now. Stay tuned to the Google Analytics blog for visibility into the feature roadmap.

Will It Be Possible to Track Visits That a Data Studio Report Received With GA4?

That feature is unavailable right now. The Data Studio release notes are the best place to go to get insight into new Data Studio features as they are released.

We Created a GA4 Property for Our App, and Created Separate GA4 Properties for Each Website We Have. Is This Following the Best Practice for GA4?

It depends on the specifics, but probably not.

Each property should correspond to a distinct user base. Then, all of your websites and apps that serve that user base should be configured as data streams flowing into a single GA4 property.

It's possible that each of your apps and websites serves distinct user bases, but we think it's unlikely to be the case based on how most organizations function.

This article elaborates on how to organize your properties and data streams

Can GA4 Replace the Custom Events Tags Created Using Google Tag Manager/Universal Analytics to Capture Specific Event-level Metrics?

We recommend parallel tracking. This means continuing to collect data in your Universal Analytics property (i.e. don't edit any of your existing UA tags in GTM) and add to them GA4 tags. You'll see a new tag template in the GTM interface. You can piggyback on triggers and variables that already exist in your container.

Once you're confident with the data in GA4, at that point you would make the GA4 property your sole source of truth and remove any UA tags.

Is There Anything We Need to Change With Setup if We Are Pushing UA in via GTM When We Create a Parallel Setup?

Yes, you need to add GA4 tags alongside your UA tags. See the answer to the previous question for more information.

Do You See GA4 Bridging the Gap Between Event-level and Session-level Metrics Eventually?

Yes, we do! They've already started to do so. Stay tuned to the Google Analytics blog for visibility into the feature roadmap.

Could GA4 Eventually Replace Adobe Analytics?

The right Analytics tool depends on several factors. This is really a decision that you would need to make.

Bounteous is a partner of both Google Analytics and Adobe Analytics. If you need help choosing between the two platforms, get in touch.

Is It Still Recommended to Deploy UA and GA4 Properties Together?

Yes, it is right now. The data model and reporting take some getting used to, and you'll be glad that you still have your comfort zone.

For companies using GA360, Google has announced they are currently in beta with an Analytics 360 version that will offer SLAs and advanced integrations with tools like BigQuery, and Google will have more to share soon. You'll want to continue with your UA properties at least until that point.

What's the Best Implementation Approach for Mobile Apps with GTM and GA?

GA4 is built using the data model from Google Analytics for Firebase and Firebase remains a very key piece of the puzzle.

First, if you don't already have a Firebase project, you'll need to create one. Within your Firebase project, you will be able to create apps. These apps appear as distinct data streams in GA4. Follow the instructions to add Firebase to your apps.

All events need to be logged from your app to Firebase. Events that you log to Firebase will appear in your GA4 property. This documentation can help you get started.

If you have other tags that need to be triggered off of your Firebase events, or if you want to be able to add/modify events without re-submitting your apps' binaries, Google Tag Manager mobile containers are the right choice for you. Add an Android container to your Android app and an iOS container to your iOS app.

Is There an Easy Way to Filter by Data Stream?

If you're thinking about filters in a UA context from the perspective that you can apply a filter and remove it from a view, then no—there are no views and consequently no filters.

If you're simply looking to see one of the standard reports to include or exclude a subset of data streams, then this is easy!

First, from any of the standard reports, click the "Customize report" button.

Customize Report button

 

Second, click on the vertical dots next to the "All Users" comparison and click Edit Comparisons. From here you can include or exclude based on the dimension Stream ID.

field showing where you can include or exclude stream IDs

Is the UA Data Layer the Best Option for Setting Up Parallel Tracking? In Other Words, Is There a Data Layer That Is Specific to GA4?

The data layer is simply a JavaScript array used by Google Tag Manager and other analytics platforms to collect various types of information. Generally, the necessary information must be placed into the data layer by the website developers.

It can be accessed by either UA, GA4, or really any other platform you're using.

We're going to assume that this question is really about the UA-specific semantics with respect to Enhanced Ecommerce projects. Google is very specific about the format of the Ecommerce information pushed to the data layer in order for UA to be able to ingest that data and turn it into reports.

Here is the UA Enhanced Ecommerce data layer documentation. You will find the updated GA4 Ecommerce data layer specifications here. GA4 is compatible with the UA Ecommerce schema, but the UA Ecommerce schema is not compatible with the GA4 schema. If you have tagged your site for GA4 Properties alongside UA, Google suggests that you do not replace your older schemas with GA4 Ecommerce data types.

Do Data Studio Reports That Are Currently Connected to GA360 Need to Be Reconfigured?

Your UA views that serve as data sources in Data Studio will continue to populate the Data Studio reports that reference them as long as you send data. If you want to use your GA4 property as a data source and report on that data, you can access it from the native connector.

data studio connector options with Google Analytics higlighted

Does GA4 Integrate With Google Ads?

Yes! We take you through the process with this blog, Linking Google Analytics 4 Properties to Google Ads: How-to Guide.

Does GA4 Integrate with Search Console and Optimize?

Not yet.

How Many Segments Can We Create in GA4?

In GA4 we create segments in the Analysis Hub. There is a limit of 200 analyses per user and 500 shared analyses per property; depending on how you configure your reports, these may serve as upper bounds.

If you're looking to use segments outside of the Analysis Hub, you need to turn them into audiences. There is a limit of 100 audiences per property.

Google publishes collection limits here.

Since One of the Unique Selling Points of GA360 Is the Integration With Google BigQuery and Given the Fact That GA4 Has an Integration With BigQuery as Well, What Would Be the Reason to Still Use GA360?

First, we would suggest that the comparison of GA360 to GA4 is the wrong comparison. Google has shared that there will be an enterprise version of GA4, so, the better comparison would be GA360 vs. the enterprise version of GA4.

Second, GA360 offers many advantages beyond just the BigQuery integration—higher data collection limits; additional features like Unsampled Reports and Data-Driven Attribution; and SLAs on data collection, data freshness, and reporting. We have several clients that do not use BigQuery and still take advantage of GA360 for these benefits.

How Will Site Search Be Impacted?

Site Search reports haven't been built in GA4 yet, but we have a walkthrough for re-creating these metrics with BigQuery that you can use in the meantime.


Make sure to check out our video session for even more GA4 answers and stay tuned to our blog for more GA4 deep-dive posts.