What is the Impact of GA4 (Google Analytics 4)?

Terence Einhorn, Sr. Solutions Consultant in Sales

Published 02/21/2024

With Google phasing out Universal Analytics in favor of their next-generation platform, Google Analytics 4 or GA4, it’s critical for marketers to understand the implications on their performance marketing practice.

Below, we will review key changes coming to GA and how marketers should (and should not) use this platform to drive growth for their business.

The'res only one true solution for media attribution in the GA4 era

Measured provides a robust incrementality capability that can be ongoingly applied to a privacy-based attribution system.

How Does GA4 Differ from Universal Analytics?

First, let's cover the distinct differences between the two platforms.

Session calculations

UA was a “session” based platform that organized data based on a particular user's session on your website. A session was defined as the period from which a user landed on your site up until:

  • User Left the Site
  • User Inactive for 30 minutes
  • The clock passed midnight
  • New campaign parameters are encountered 

In GA4, the first two will still time out a session, but the latter two will not, leading to more complete and robust tracking across your site.

Mobile App and Website Tracking

Unlike UA, which required setting up a new property to track App-based events, GA4 allows brands to track Web and App events alongside one another in the same dashboard.

As a result, marketers will have a much more comprehensive product experience and richer analytics that cover a broader swath of their business.

Cookies and IP-less tracking

Perhaps the most anticipated of key changes, GA4 will phase out the use of third-party cookies and IP address collection in an effort to make the platform as privacy-centric as possible.

GA4 will instead use first-party cookies, supplemented with AI algorithms that will help “fill in the gaps” where cross-platform tracking is missed.

Machine Learning

On that note, GA4 leverages AI/Machine Learning for a number of applications.

Namely,  it will use signals from across the platform to provide predictions for Purchase, Churn, and Revenue probability for a given source.

Customizable Dashboards

GA4 will offer a host of customization optimizations to Dashboarding and reports that were not available in UA.

Additionally, they will enable seamless (and free) Bitquery integration moving forward.

How Do I Identify Key Limitations in GA4’s Approach?

Now, let’s take a look at the key limitations of GA4 that marketers will need to be aware of before using it as a one-stop shop for performance marketing optimization.

GDPR Compliance

For starters, this privacy-focused platform is not actually privacy-compliant everywhere. GA4 has been banned in the following countries:

  • Austria
  • France
  • Italy
  • Netherlands
  • Denmark

Data Sampling

GA4 has increased its reliance on sampling of user data to project the behavior of an entire audience or market.

While this is a necessity in the privacy era, the quality and consistency of that sampling on GA4 is highly opaque, meaning it’s difficult to gauge the accuracy or reliability of sampled data.

Removal of Attribution Models

UA allowed marketers to select a variety of different models as the basis for attribution. However, GA4 will sunset all of these options except for two, citing the difficulty that “rule-based” models have in accurately tracking conversions.

The two remaining models will be:

Data-Driven Attribution (DDA) 

Data-driven attribution will be the default model in GA4 and relies on historical data from within your account to estimate the impact of a given media source on conversion behavior. While this is a more advanced technique than “Last Touch,” there is a considerable “black box” that doesn’t allow marketers to see the various inputs going into the model and how these inputs are impacting resulting attribution, making it potentially difficult to fully interpret results.

Last Touch (LT)

Last Touch tracking is the “old faithful” of marketing attribution but remains a highly flawed practice.

Due to the same tracking limitations that plague the other rules-based models, attributing sales to media based on the last click a user made before conversion has a few critical issues:

  • View Through: Like other GA attribution models, Last Touch cannot identify users who viewed a Social or Video ad and then proceeded to buy a product. The conversions erroneously get attributed to direct or organic sources.
  • Lower Funnel Bias: Most Search, especially Brand Search “conversions,” are preceded by another media in the funnel. However, Last Touch modeling does not take any of these exposures into account, leading to highly inflated performance estimates for lower funnel tactics like Search or Affiliate marketing.

Incrementality

The most important limitation of both DDA and Last Touch is that they are unable to take into account incrementality, or in other words, they fail to answer the question: “How many of these sales would have happened anyway, even without the marketing exposure?”.

This is a critical piece of any marketing attribution component and is fundamentally missing in any tracking-based solution such as GA4.

For a full rundown on how to measure and account for incrementality in marketing measurement, check out this FAQ.

Is Google Analytics Even Relevant Anymore?

Yes and No.

Google Analytics is still a relevant tool for certain analyses, such as site optimization and high-level funnel reporting. It also integrates seamlessly with Google Ads/Google Tag Manager to enable more sophisticated campaign optimization across Google properties.

However, brands using GA4 as a marketing attribution tool risk making very costly mistakes. Even before cookie depreciation, using any site-side analytics tools as a means of tracking marketing performance was highly problematic:

For one, almost no “view through” media has ever been measured by GA, meaning a user has to physically click on a social or video ad in order for that ad to be credited with an eventual conversion. This leads to a vast under-crediting of view-based platforms and their contribution to your business.

Furthermore, GA4 (and Google Ads) have an incrementality problem, driven by their use of AI in targeting and optimization. 

Their aim is to target users with the highest predisposed propensity to purchase your product to make their conversion rates look as good as possible.

The obvious problem with this is that this aims ads at users who are already very likely to convert before you serve them an ad, meaning a large portion of this budget will be inherently redundant.

Without being able to quantify how many of these conversions “would have happened anyway,” it is impossible to accurately evaluate a media’s true impact on the business.

Overcoming GA4 Limitations: A Mixed Approach to Marketing Attribution 

As a result, the only true solution for media attribution in the GA4 era is a robust incrementality capability, ongoingly applied to a privacy-based attribution system.

To read more about how Measured tackles this problem, see here