Comparing Universal Analytics (UA) and Google Analytics 4 (GA4)
Written By: Shane Clark on June 12, 2024
As Google started to warn the world of the pending demise of Universal Analytics (UA) to be replaced with Google Analytics 4 (GA4), I figured this was an excellent opportunity to compare both platforms. The current “threat” from Google is a total loss of data by the end of 2023 without migration to the GA4 platform. Therefore, this blog will cover the fundamental differences between the two platforms.
With GA 4, Google Analytics has received a most notable, major update. GA4 can now track users across various devices, domains, apps, platforms, and throughout several sessions. This gives you a better idea of how your customers engage with your business and brands.
Google Analytics 4 is an entirely new property. Analytics App + Web was the previous beta version. GA4 can automatically offer beneficial insights thanks to machine learning (AI), providing a better and richer understanding of your clients across devices and platforms.
Like earlier versions of Google Analytics, GA4 is not backward compatible or an “out of the box” solution. The platform has, however, been enhanced significantly. New pathing components, funnel tools, and BigQuery are all new to GA4.
The What vs. Why Models
From the highest level, I see the differences, and the direction that Google is going is into more of a “why” model. UA offered us a more accessible “what,” meaning what is happening on the website/app, but did not always allow for the “why.” As I explain some of the event-driven modeling parts of GA4, this will make more sense. The “why” offered as part of GA4 is “why is this happening?,” allowing marketers to understand user and data acquisition modeling to make changes that can improve conversion rates.
Google Analytics 4 (GA4) Is Event-Driven
To leverage the power of GA4, it is vital to understand how events work. If you look at the explanation from Google support, you can see that the traditional page views, social, transactions/e-commerce, user timing, exceptions, and app/screen views have all been replaced with events. These events are the foundation of GA4.
Understanding the Importance of Google Tag Manager (GTM)
Google Tag Manager will be vital for any successful implementation of Google Analytics 4. The GA4 system relies heavily on events fired from GTM to be able to read elements and user data. Even with UA, you could still read most of your data and create a practical implementation to read, store and manage the data/metrics you required.
Options to End a Session in Universal Analytics (UA)
1 – The first option is the session will expire after thirty minutes of user inactivity or at midnight
2 – The second option is when a user comes by way of a campaign, leaves, and returns through a different campaign.
Options to End a Session in Google Analytics 4 (GA4)
In GA4, a session only ends after thirty minutes of user inactivity. Because of the way sessions are measured, you most likely will find few sessions being tracked in your GA4 reporting over the same period being measured compared to UA.
Ability to Track Users Across Multiple Platforms/Devices
GA4 allows for tracking users’ data from a website and app. In Google Analytics 4, a new feature called Attribution was released. Enhancements to attribution in GA4 properties include a new conversion pathways report and additional attribution features such as property-level attribution modeling.
This enables a more in-depth examination of conversions and which points along the conversion path assist customers in completing a conversion event.
Attribution modeling can help with this. Other searches, ad clicks, and other points along the customer journey can be credited for site conversions using attribution. An attribution model is a rule, a series of rules, or a data-driven algorithm that decides how conversion credit is distributed across conversion channels.
Ability to Detect Data Anomalies More Effectively
Analytics Intelligence employs anomaly detection, a statistical approach, to find anomalies in time-series data for a specific measure and anomalies within a segment at the same time.
Analytics Intelligence forecasts the metric’s value for the current period using past data and marks the datapoint as an anomaly. This happens when the actual value falls outside the credible interval. The training duration for detecting hourly abnormalities is two weeks. A training period for detecting everyday abnormalities is ninety days. The training duration for detecting weekly abnormalities is about thirty-two weeks.
While time-series anomaly detection employs historical data to identify a single metric inside a single dimension value, we also offer anomaly detection across several metrics and dimension values at a single moment in time.
GA4 Offers Up to 50 Data Streams
There are up to 50 data streams per Google Analytics 4 property (any combination of web data and app streams, including the limitation of 30 app data streams). A data stream is a data flow from a client touchpoint (such as an app or a website) to Analytics.
Analytics provides a snippet of code that you can add to your app or website to gather data when you create a data stream. Data is collected from the moment the code is added, and this data is used to create your reports.
You build a data stream for each platform if you’re gathering data for a single logical application on numerous platforms like iOS and Android (note that’s one data stream for Google Play and another stream for the app on App Store). If you delete a data stream, Analytics keeps the previous data, but you won’t be able to analyze it anymore or use it in report filters.
GA4 Analytics Insights
This is a set of features that leverages machine learning and parameters you establish to help you better comprehend and act on current data.
Analytics Intelligence automatically recognizes odd changes or emerging trends in your data and tells you on the Insights dashboard within the Analytics platform.
You define conditions that detect significant changes in your data. When the circumstances are met, the insights appear on the Insights dashboard, with the option of receiving email alerts. Per property, you can generate up to 50 unique insights.
This Insights dashboard displays your property’s most recent automated and personalized insights. As new insights are triggered and Analytics discovers intriguing insights, the dashboard is updated in real-time. To see more information on the cards, click on them. To access prior understandings, click Load more. After they are generated, insights are kept for a year. Analytics Intelligence learns the insights you’re most interested in after each interaction and ranks your new insights accordingly.
Sessions Are Computed Differently in GA4
Google Analytics 4 has some differences from Universal Analytics in session counting. A new campaign in Universal Analytics will start a new session regardless of activity. A new campaign does not create a new session in Google Analytics 4. This could result in decreased session counts in Google Analytics 4.
Late hits could also play a role. Hits that aren’t sent right away are known as late hits. Hits are processed in Universal Analytics if they arrive within 4 hours of the previous day’s closure. Events are processed in Google Analytics 4 if they arrive up to 72 hours late.
You may see larger session counts in your GA4 property and variances in reported data because Google Analytics 4 events are handled over a longer time.
For example, a person loses service while accessing your website on their mobile device, only to regain it 48 hours later.
Google Analytics 4 processes the late hit, whereas Universal Analytics does not, resulting in a greater session count in Google Analytics 4.
When iOS apps are backgrounded, Google Analytics 4 events are immediately submitted. In UA, this is not the case. As a result, iOS-related metrics in your GA 4 reports may be much higher.
How GA4 Complies with Privacy Regulations
GA4 anonymizes all users’ IP addresses by default. This option cannot be changed. In comparison to Universal Analytics, which by default monitored IP addresses, breaking GDPR guidelines that considered an IP address personally identifiable information (PII).
Google Analytics 4 cannot control where data is stored, just like Universal. Because most GA4 servers are in the United States if you’re in the European Union and want to comply with GDPR, ensure that your data privacy strategy includes a statement that international data transfers will occur.
GA4 collects data on users as they engage with your website or app, like Universal Analytics, and many privacy laws, including GDPR, require users to opt-in to this data collecting. Therefore, your cookie banner should explicitly indicate what tracking the user is opting in/out of and provide clear opt-in/out options to obtain users’ consent.
The Permission Mode function in Google Tag Manager was announced late in 2020 and is still officially in “beta,” it allows you to customize your Google tags (Analytics and Ads) to respect users’ consent choices.
When you start a new Google Analytics 4 implementation, you may utilize Consent Mode to create your GA4 tags right away, ensuring that your tracking responds appropriately to users’ opt-in/out decisions.
GA4’s Attribution Modeling
The act of allocating credit for conversions to different adverts, clicks, and elements along a user’s conversion path is known as attribution. An attribution model is a rule, a series of rules, or a data-driven algorithm that decides how conversion credit is distributed across conversion channels.
Cross-channel rules-based models, an Ads-preferred rules-based model, and data-driven attribution are the three attribution models offered in the Attribution reports in Google Analytics 4 properties.
Improved attribution tools, such as a revised conversion pathways report, and new attribution features, such as property-level attribution modeling, are available in Google Analytics 4 properties, providing deeper insights and more actionability than ever before. For example, before making a purchase or completing another valuable activity on your website, customers may conduct multiple searches and click on several of your advertisements. Typically, the last ad customers click is given full credit for the conversion.
GA4’s Machine Learning Analytics
With machine-learning-powered predictive analytics, Google Analytics 4 promises to predict consumers’ future behavior. In addition, this feature can supposedly increase model quality by using shared aggregated and anonymous data.
The feature works only if at least 1,000 returning customers triggered the appropriate prediction condition during seven days. It also won’t operate if the model isn’t maintained over a period.
In the event that the model quality falls below the minimum threshold, Analytics will stop updating the corresponding forecasts. As a result, most analytics users are unlikely to benefit from GA4’s machine learning insights.
Behavior Flow Report’s Been Replaced
The Behavior Flow report in Universal Analytics allows you to see how visitors move from one page or event to the next. It comes in handy when you need rapid and straightforward information. However, it is no longer available in Google Analytics 4, and it has been replaced by two new overly burdensome reports: the funnel exploration report and the path exploration report.
GA4 Custom Dimensions
To capture complex data, you can use Google Analytics 4 to construct custom dimensions. For example, if someone reads a page on your site, you may add custom dimensions like author name or blog post length to the data. However, you may only use up to 50, rendering features like this essentially useless for others.
Views
The ability to configure views is a crucial feature of Universal Analytics. Views enable you to create specific analytics settings for testing or data cleansing, such as filtering out internal traffic. Views are useful for sifting data quickly and simply. Smaller organizations, casual consumers, and do-it-yourself marketing departments will benefit from preset ideas that contain only the information they need to see.
User ID and Client ID Functionality
User ID in UA and GA4 aims to provide an identification space for users to evaluate their data. To translate user IDs in a Universal Analytics property to a Google Analytics 4 property, no additional adjustments are required in terms of data collecting. The User ID property in Google Analytics 4 shows how users interact with your app or website across platforms and devices.
You must be able to produce your own unique, persistent IDs, assign those IDs to your users consistently, and include the IDs with the data you provide to Analytics to use this functionality.
From all of the data associated with the same user ID, Analytics builds a single user journey. A Google Analytics 4 property, unlike Universal Analytics, uses the native integration of User ID across all reporting, analysis, and insights and does not require a separate User-ID reporting view.
A Client ID is a one-of-a-kind, randomly generated string that serves as a pseudo-anonymous identifier for a browser instance. It is saved in the browser’s cookies, allowing subsequent visits to the same site to be linked to the same user.
The semantics of Client ID in Universal Analytics and Google Analytics 4 are similar, and they both fulfill the same objective of giving a pseudonymous user identity. The App Instance ID is the Google Analytics 4 property counterpart for apps.
Audiences Have Changed
Audiences allow you to classify your consumers in valuable ways for your company. To include nearly any subset of users, you can segment by dimensions, metrics, and events. New information about users may become available. When this happens, the system reevaluates their audience memberships to verify that they continue to fit the audience criteria. They are removed from those audiences if the most recent data indicates they no longer satisfy the requirements.
Suppose you link your Analytics account to Google Ads and use the default setting of Enable Personalized Advertising. In that case, your audiences will appear in your Google Ads shared library, where you may use them in your ad campaigns. You can remarket to prior or existing consumers and generate similar audiences to find new people. When data is available, the system prepopulates the user lists in Ads that match your audiences with up to 30 days of data.
You can find the same user groups based on your Google Analytics 4 data and export those audiences to your linked Google Ads accounts to solve the same use cases by migrating audiences from your Universal Analytics property to your Google Analytics 4 property.
Per Google Analytics 4 property, you can build up to 100 audiences. Unlike Universal Analytics audiences, audiences in Google Analytics 4 properties update everywhere instead of needing you to choose a destination.
If you need more audiences:
– Combine Universal Analytics audiences in your Google Analytics 4 property
– Upgrade from Google Analytics 4 to Google Analytics 360, which has a 400-audience maximum.
– Google Analytics 4 doesn’t support some Universal Analytics dimensions.
Audience Types and Triggers
You choose the industry category for your business when you create a property. Then, analytics presents several preset proposed audiences based on your choices based on the recommended events for that category. As a result, the same audience can show up in various industry sectors.
Although you will only see options relevant to your category, you can build any of these audiences if they are relevant to your business. But, first, you’ll need to gather the events and criteria that make up the audiences you’ll be using.
You can change these proposed audiences as needed.
Click Audiences on the left, then New audience to see the proposed audiences for your property.
Preconfigured Audiences in GA4:
Top players
Users who have attained a certain level.
Top scorers
Users who received a high rating.
Tutorial abandoners
Users who did not finish the tutorial.
Tutorial finishers
Users who have finished a tutorial.
Video completed
Users who have completed a video.
Video start
Users who have started watching a video.
Wishlist users
Users who made wishlist additions.
Item searchers
Users who conducted item searches
Item viewers
Users who looked at specific products
Leads
Users who have the potential to become business leads.
Searchers
Users who have conducted any type of search
Streamers
Users who watched stuff online.
Achievers
Users who have completed a specific goal.
Billable users
Users who registered a payment method.
Cart abandoners
Users placed things in their shopping carts but did not complete the transaction.
Checkout starters
Customers who started the checkout process but did not complete it.
Audience triggers allow you to set off events when people meet the criteria for joining an audience. For example, you can set up events when users hit certain milestones, such as reading a certain number of articles, starting sessions, or when they cross convert criteria.
You can analyze these events in reports. You can also enable them as conversions just like any other event. A single event can meet all the audience requirements. When this happens, the system replicates metadata from that event to the triggered event.
Multiple events can satisfy audience requirements. In this case, the system replicates metadata from the most recent event to the triggered event. The system does not include the value associated with the event in the copied metadata. The metadata contains the timestamp, session, and screen/page information.
GA4 and Google Signals
Google signals are site and app session data that Google identifies with users who have signed into their Google accounts and enabled Ads Personalization. This data association enables cross-device reporting, cross-device retargeting, and cross-device conversion export to Ads for these signed-in users.
Google Analytics 4 exports anonymized cookie-related event data to BigQuery. This may theoretically lead to the system counting same user numerous times on different devices. Google Signals, on the other hand, tracks users across devices. Compared to reports based on Google signals data, data uploaded to BigQuery may reveal more people.
More on Cross-Platform Reports
Using your User-ID or Google-signals data, connect data about devices and activities from different sessions to better understand user behavior at each stage of the conversion process, from initial contact to conversion and beyond.
The system requires a monthly average of 500 users daily for your reports to include Google signals data. With the addition of Google signals data, Google Analytics increases the accuracy of user counts in reporting. In addition, it evaluates users who do not have a user ID using cross-platform audience criteria.
Remarketing
Create remarketing audiences based on your Google Analytics data and share them with your associated advertising accounts.
When you enable Google signals, Google Ads and other Google Marketing Platform advertising solutions can display ads in eligible Cross Device-campaigns.
Ad Reporting
Google Analytics gathers data based on your tagging preferences, Google signals data, and Google advertising cookies.
Interests and Demographics
Google Analytics collects additional demographic and interest data. The system collects this data from both device identifiers and users. These devices and users must have signed in to their Google accounts. The devices and users must also have switched on Ads Personalization. However, you won’t be able to use these services individually if you disable Google signals. The system will disable all of these capabilities when you disable Google signals.
Ad Personalization
When consumers enable Ads Personalization, Google can build a comprehensive picture of how they engage with an online business across multiple browsers and devices. For example, you can observe how consumers browse things on your site on their phones, then return later to finalize purchases on a tablet or laptop.
Google signed-in data will expire in twenty-six months. This is a default setting. If you set the Analytics Data Retention to less than 26 months, the Google signed-in data will follow suit. The system only includes aggregated data in the Cross-Platform reports. The system will never disclose individual user data.
Remarketing in GA4 and Google Ads 360
You can use the Google Analytics 4 audience builder, which has comprehensive list-building features, to construct remarketing audiences. For example, you may use Google Ads to personalize the campaign for people who have previously visited your site or app by using these audiences. In addition, Analytics provides detailed user analytics that might aid in creating remarketing lists.
When you link your Analytics property to Google Advertising, you must activate Google signals and allow ad personalization to use a Google Analytics 4 audience with your Google Ads campaigns. When you build an audience in Google Analytics, it will automatically appear in Google Ads, where you can link it to your search ad groups.
Google associates visit your site or app (based on your Analytics criteria) with one of Google’s advertising cookies on users’ devices. When your customers later search on Google.com, they may see customized ads based on their previous interaction with your business.
Remember the following:
– A remarketing list for Google search advertising has a maximum lifespan of 540 days.
– Before you can use a remarketing list for Google search ads to customize your search ads, it must have at least 1,000 cookies. This helps to safeguard the privacy of persons on your list.
– Google has included Google Display Network demographics dimensions in remarketing lists. However, RLSA does not consider age, gender, or interests.
– Remarketing audiences are available in Google Ads and Display & Video 360. All Analytics accounts can use the Google Ads integration. However, only Google Analytics 360 accounts have access to the Display & Video 360 integration.
– You specify the audience parameters and the advertising accounts in which you wish to employ the audience to establish a remarketing audience.
You can select the audience criterion by:
– Choosing from a library of audience definitions.
– Make a new audience category.
– Importing a section
When you store an audience, it becomes available in the advertising accounts you specify, where you can use it in your remarketing campaigns. You must add an audience to at least one of your ad groups in Google Ads to use it.
Can You Turn Off Ads Personalization in GA4?
On an ongoing basis, you can accept or prevent the usage of analytics data acquired from users depending on their country and area for ad personalization. If you change the advertising customization setting, it does not affect data gathered for Google signals, regardless of whether you have enabled Google signals. You can turn off advertising personalization for a country or region. The system will collect Google signals data from users in that area for measurement purposes only and not use it for personal advertisement.
Google Ads, GA4, and Current Analytics
Connect your Google Ads account to your Analytics property to observe the entire customer journey, from how customers interact with your marketing (for example, by clicking ads) to how they complete the goals you’ve established on your site or app (e.g., making purchases, consuming content). You can utilize the import tool instead of manually linking many Google Ads links if you’re switching from Universal Analytics to Google Analytics 4 and need to migrate multiple Google Ads links.
You can do the following when you link your Google Analytics 4 property to your Google Ads account:
– In the Acquisition Overview report, you can see your Google Ads campaigns.
– In the User Acquisition report, you can see new Google Ads dimensions.
– Enhance your Google Ads remarketing with Analytics audience data by importing Analytics conversions into your account.
– In the Advertising workspace, you can see your Google Ads campaigns, including the Attribution reports.
Configuration
An editor can link individual Google Ads accounts and management accounts to Google Analytics 4 properties. Each property can have up to 400 links. If your existing Google Ads configuration exceeds this limit, consider setting up a Google Ads management account and connecting it to your Analytics property. A Google Ads management account link counts as one link.
Use a Google account with the necessary permissions to link a property to Google Ads. Ask your Analytics or Google Ads administrator for assistance if your account does not have these rights.
You must have the Editor role for the property you want to link in Analytics. The system requires that same Google account for administrative access in Google Ads. Any data you import from Analytics will be available to all your client accounts if you link to a Google Ads manager account.
A user can unlink the associated Google Ads account. A Google Ads Admin or an Analytics user with the Editor authority must do this. Data between the associated Google Ads account and the Analytics property stops flowing when you delete a link.
The system does not display unlinked Google Ads account data (e.g., clicks, impressions, CPC, etc.) in Analytics reports. However, historical Google Ads dimensions (e.g., Campaign name, Ad group ID) will be available in Analytics after removing the link. In addition, new data for these dimensions coming from clicks from an unlinked Google Ads account and Google Ads metrics (e.g., Cost, Clicks) for all date ranges will appear as (not set).
The system no longer includes new users in unlinked Google Ads accounts in Analytics remarketing audiences. In addition, the system no longer imports conversions from Analytics when a user unlinks Google Ads accounts.
GA4 Advertising Snapshot
The Advertising Snapshot report allows you to rapidly review your business stats before diving further into the areas you want to investigate.
The Advertising Snapshot includes data from June 14, 2021, onwards in the summary cards in this overview report. Each card answers a single question about your company: What is the average time it takes for clients to convert? To see more detailed information, click the report name at the bottom of the card. Default channel grouping conversions discover which of the default channel grouping’s channels resulted in the most significant modifications.
This card uses the reporting attribution methodology to credit conversions to specific channels.
Conversion paths show conversion paths that led to the most conversions for the time period you chose.
Model comparison examines how the system distributes credit across channels using various attribution models. To compare multiple attribution models, use the drop-down menus at the top of the card.
Summary:
Google Analytics 4’s event-driven model emphasizes the “why” more than just “what”. This will undoubtedly change how everyone involved will do digital marketing. If you are migrating from Universal Analytics to Google Analytics 4, you can contact Shane today for help.
