Google Analytics 4 Migration Guide

Making the switch to GA4 involves a learning curve. We've been using the same reports for years, and now many metrics and reports we rely on are no longer available in GA4.

With this guide, we hope to shed some light on what you can expect with the migration to GA4 and offer our help rebuilding your web analytic practice.

If all of this makes your eyes glaze over, contact us at XD Lab, and we will do the setup for you.

It's All Events

GA4 has a new and entirely different data structure than Universal Analytics. Instead of recording sessions and pageviews, the GA4 model is based exclusively on events.

For user experience designers like us, bounce rate was a critical signal that a page was underperforming due to the page design or a traffic source mismatch. Without that functionality, GA generally becomes less practical for easy use. And if bounce rate was one of the critical metrics for your business or your clients, you'll need to find a new way to identify your best and worst performing pages. There are exciting ways to leverage event data to determine overall page performance, but it will require setup and changing how we think about web analytics.

GA4 provides metrics for "Engaged sessions," "Average engagement time per session," and "Engaged Rate." But be careful! It would help to rely on external data analysis to get actual session counts. Google employs the approximate calculation method (HyperLogLog++ algo), which is an estimated calculation of the total sessions on your site.

The data structure of a session (necessary for the custom SQL reporting we will most certainly need to do) is as follows:

When a session starts, Google automatically collects a session_start event and generates a session ID (ga_session_id) and session number (ga_session_number) via the session_start event.

• Session ID identifies the session that an event came from. For example, two different session IDs are generated when a user has two sessions on your site.
• Session number identifies the number of sessions that a user has started up to the current session (e.g., a user's third or fifth session on your site).

You can expect your total number of sessions reported in Google Analytics to decrease once the GA4 migration is complete.

Some metrics are different, but many regular reports from Universal Analytics are unavailable in GA4.

Where Are My Reports?

Say goodbye to the long list of reports in the side panel. GA4 has a limited number of prebuilt reports. It helps to think of GA4 as a data collection engine with much of the analysis performed outside the platform rather than a single pane of glass.

The GA4 reports that exist are limited and sometimes need to be clarified with the removal of report-friendly data labels. You may find narrowing down complete URLs in landing page reports and paths in referral traffic difficult. Many users, including our work at XD Lab, built their GA4 reports in 3rd party data analysis tools or exported data for further SQL-based analysis.

What About Conversions?

The data model in GA4 is not based on sessions and pageviews. That means that 'thank you page' conversions no longer exist. Calculated metrics like Pages / Session and Session Duration don't exist in GA4, and a destination goal based on a view of a specific page is now a complex endeavor. Losing easy Destination goals like Pardot or Marketo 'thank you pages' will drastically change how many B2B companies track conversion metrics. Since data in GA4 is processed as various types of events, we only have conversions based on event data.

If you rely on thank you pages to record events and conversions, you must update your analytic tagging as soon as possible.

If you are currently using embedded 3rd party iframe forms (like Pardot, Marketo, or Eloqua forms), GA4 form interactions will not work out of the box. Similarly, forms submitted in the background using Javascript are not captured and will require custom events, dimensions, and metrics.

Best practices for conversion tracking with GA4 will include enhanced measurement for form submissions. Enabling advanced tracking on form interactions will provide the following events:

• 'form_start': the first time a user interacts with a form in a session
• 'form_submit': when the user submits a form

For our GA4 property, we've enabled the additional parameters:

• form_id – ID attribute of form DOM element
• form_name – Name attribute of form DOM element
• form_destination – Action attribute of form DOM element (where form data is sent)
• form_submit_text – Text of the submit button, if any. (only form_submit event)

These parameters are not available for reporting by default. It would be best to create custom dimensions to make them available in custom reports based on these values.

What About Historic Data?

Any data collected before the July deadline will most likely be lost. Google strongly encourages you to export your historical reports asap.

You must export and house the data independently, preferably in a queryable database with lightweight data visualization tools.

If you want assistance setting up a data archive and reporting portal, we at XD Lab can spin up your environment to preserve your website's data.

XD Lab creates integrated marketing platforms. Technical disconnects between systems and subsequent data loss only result in waste and operational inefficiencies. Our superpower is being able to effortlessly move from the pixel to the person. This forest and the trees approach brings efficiency and effectiveness to the marketing programs we run.

If you'd like to get started with a GA4 migration, talk to us today.