back to thedrum.com
The marketing sector would not exist as it does today were it not for third-party cookies. Over the past two decades, businesses have developed a heavy reliance on these bite-sized chunks of data for user insights, audience analysis and strategy development.
However, this method has faced frequent controversies and raised countless concerns regarding user privacy. As a result, the digital hemisphere is shifting, with tighter GDPR regulations leading to a future free from third-party cookies; by the end of 2024, this method of data collection may be a thing of the past.
In light of this change, alternative methods of data collection, prediction and analysis must be harnessed, filling the gap left by cookies. Machine learning algorithms, particularly those within Google Analytics 4, will play a significant role in compensating for this loss.
Before exploring the key components of this so-called ‘cookieless’ future, it is important to clarify the difference between first and third-party cookies. First-party cookies are accessible only by a website’s owner and play a vital role in collecting analytical data and optimising site functionality. In contrast, third-party cookies, created by separate domains, track a user’s behaviour across the internet. In this context, ‘cookieless’ refers to eliminating third-party tracking by internet service providers (ISPs).
Eliminating third-party cookies is certain to leave voids in online marketing. Google heavily relies on cookies for its advertising platform. However, to ensure privacy compliance, alternative methods must fill these gaps; tools such as statistical modelling, predictive analytics, and machine learning will play critical roles.
As of July 2023, Google permanently retired Universal Analytics, replacing it with Google Analytics 4 (GA4), a property now used by over 20 million websites worldwide. This transition brings significant changes for digital marketers. With a focus on user privacy, GA4 utilises AI algorithms to connect data points without third-party cookies.
By blending directly observed and modelled data, GA4 builds comprehensive and accurate datasets while complying with user privacy regulations. Moreover, AI integration and machine learning enable more precise data collection than third-party cookies, utilising predictive analytics, analytics intelligence, and behaviour modelling to compensate for cookieless browsing data loss. Below, we delve into the specifics of these features.
Predictive Analytics
Among GA4’s many innovative features is the property’s capacity to predict future metrics and audiences through machine learning algorithms. As soon as data collection begins, GA4’s algorithms start learning from the information gathered, enabling the generation of projections like revenue, events, and trends based on account-specific datasets.
Additionally, GA4 utilises the data from its machine learning algorithm to create predictive audiences, forecasting which of your audience is likely to churn based on past events. This information is set to prove infinitely valuable in the development of remarketing strategies.
Consent Mode & Behavioural Modelling
Moreover, GA4’s ‘consent mode’ is vital for data compliance. This feature allows businesses to create predictive data even when a user doesn’t consent to data collection. When consent is declined, GA4 uses relevant existing data to predict the user’s behaviour.
This method of behavioural modelling fills data gaps when direct observations are lacking or non-existent. Instead of cookie-shaped voids, GA4 combines observed data with behaviour models, offering comprehensive insights into user journeys and site interactions.
Analytics Intelligence
Alongside the above, GA4’s Analytics Intelligence functionality is set to play an important part in navigating a cookieless future. As Google describes it, analytics intelligence uses machine learning and custom configurations to understand and act on data effectively. In other words, machine learning can identify anomalies and errors while simultaneously adapting to these variations; as GA4 receives more distinctive data, its ability to detect outliers improves.
This advancement in GA4 distinguishes authentic data from anomalies, reducing reliance on third-party cookies for true ‘insight’. Unlike cookies, which are about 60% accurate at best, machine learning and analytics intelligence offer continuous enhancement opportunities.
As previously mentioned, past iterations of Google Analytics faced scrutiny for their handling and retention of data. GDPR regulations have, naturally, tightened over the past few years; Google has adjusted its software accordingly.
Alongside the implementation of machine learning algorithms, GA4 introduces a host of features focused on data compliance and user privacy. Most notably, IP anonymisation means that IP addresses are neither logged nor stored. In the tech giant’s own words, ‘analytics drops any IP addresses that it collects from EU users before logging that data via EU domains and servers.’ Whereas universal analytics collected IP addresses by default, this is no longer possible in GA4. Regarding EU privacy legislation, this is perhaps the most momentous update.
Of course, this fortification of user privacy is just the tip of the iceberg. Alongside the above mentioned IP anonymisation and consent mode, Google have introduced:
Google Analytics exists as the dominant property for good reason; its newfound capabilities balance cutting edge technologies with tightened data restrictions. That said, GDPR legislation differs in relation to server location. To ensure compliance in your region, it is essential that your property is set up in a way that does not breach privacy laws.
Without a doubt, the accessibility of machine learning in GA4 is an exciting development. Unlike previous instances that required third-party solutions, Google’s integrated features now place AI power in the palm of marketers’ hands.
Machine learning, especially that within GA4, will be crucial in filling the gaps left by third-party cookies. While the property is far from flawless, businesses should not underestimate the potential of machine learning in this platform. In a cookieless future, a certain adaptability and open-mindedness towards technology is essential.
