GA4 vs UA: Breaking Down the New Terminology
Google Analytics 4 is the next-generation measurement solution. Come 1st July 2023, standard Universal Analytics properties will stop processing data. But what are the key differences between GA4 and UA?
Measurement
UA was a session & hit-based measurement model built primarily for tracking interactions within a limited time domain from a single browser. Meanwhile, GA4 is user & event based designed to track interactions across multiple devices / sessions.
Attribution Modelling
Rather than being almost entirely last-click attribution with a bit of buried attribution modelling, GA4 brings different attribution models front and centre. Depending on the context you might be using first click, last click, or data-driven.
This sounds complex but it is about giving you the right tool for the job. Last click is terrible at identifying how you acquire customers, first click has limited value when trying to determine which clicks actually convert.
Customisation
Because of the more versatile measurement model, you can make it fit a far wider range of online businesses. UA was effective for content websites and ecommerce but what if you had a single-page web app or an online game? GA4 makes it far easier to build a measurement plan to suit your business.
An extension of that is GA4 enables you to customise the interface so you can hide any reports you don't use or even create new ones.
Reporting
Then there is the explore section (custom reports). We only used to have funnels for Shopping or Checkout behaviour, now you can build them for any sequence of interactions even if they happened across multiple sessions.
Want to build a funnel to see how many people read a certain bit of content, subscribe to your email lists, come back from a future email campaign, and ultimately submit a lead form? You can now.
Want to find which channel acquires the customers with the highest LTV? It's right there. The Explore section really puts the "Analysis" back in Analytics.
Audience Building
Segments are now called Comparisons for some reason, but they are basically the same thing.
Audiences are the close cousin of Comparisons - set certain behaviours or attributes and when a user meets the criteria they are added to that audience. You can then analyse how people in those audiences behave differently to those in other audiences (e.g. how do your purchasers behave compared to non-purchasers) and even use them in your Google Ads campaigns.
And busy enough sites will give you predictive audiences based on machine learning, for example, your audience of most likely to purchase in the next week. Powerful stuff!