The transition from Universal Analytics to Google Analytics 4 is complete, marking the end of an era for businesses, marketers and data analysts worldwide. Following this change, it’s only right that we should question its capabilities.
What is Google’s next-generation data property really capable of? How will it change the ways in which we record and utilise data?
Most importantly, can it really predict the future?
Of the many innovative features present within GA4, its predictive capabilities and integrated machine learning algorithms are the most striking. In the right hands, these tools offer enriched data collection and advanced insights for businesses, though there are certain limitations.
Of course it’s important to remember that predicting the future is impossible, even for Google. That said, GA4’s machine learning algorithms and predictive analytics features bring us closer than ever. Using data collected from past user behaviour, the property is able to predict which customers are likely to convert, channels that will lead them there, and the devices they are most likely to browse from.
It is worth noting that these predictions are limited by both the accuracy and totality of the data from which they are drawing. In other words, the better quality the input you give GA4, the more accurate the output.
As it stands, GA4 is able to generate predictions based on three primary metrics: purchase probability, churn probability and predictive revenue.
This metric refers to the probability that a user who was active in the last 28 days will log a specific conversion event within the next 7 days.
As the name suggests, this metric refers to the probability that a user who was active on your app or site within the last 7 days and will not be active within the next 7 days.
Predicted revenue forecasts total revenue from all purchase conversions within the next 28 days from a user who was active in the last 28 days.
As touched upon earlier, future metrics are driven by predictive modelling. What this means, is that by training machine learning algorithms to respond to certain datasets, GA4 is able to forecast activity, behaviour and trends. This information can then be used to create predictive audiences.
In Google’s own words, a ‘predictive audience is an audience with at least one condition based on a predictive metric. For example, you could build an audience for ‘likely 7-day purchasers’ that includes users who are likely to make a purchase in the next 7 days.’
There are a multitude of scenarios in which predictive audiences can be used to enrich data and engage with potential customers, such as remarketing or re-engagement campaigns.
As is the case with all machine learning tools, a certain amount of pre-existing data must be collected in order to generate predictions.
Firstly, you’ll need to have at least 1000 positive and negative samples. In other words, GA4 requires data on at least 1000 purchasers or conversions (positive) and 1000 users who have not purchased (negative). Additionally, the number of non-operative or inactive users on your site must be at least 1000.
Moreover, GA4 requires the above mentioned model quality to be maintained for a certain amount of time. Usually, this is around 28 days. Audiences that do not meet these prerequisites will be deemed ‘not eligible to use’ until sufficient data is gathered.
For those working within data-driven marketing, predictive analytics exist as an opportunity to leverage insights and increase the effectiveness of your campaigns. Whether you’re looking to reduce churn rates, optimise campaign performance or ready yourself for the fast-approaching cookieless future, there are a plethora of ways in which businesses can use predictive analytics to their advantage.
As a data driven SEO agency, we make it one of our primary missions to maintain a comprehensive and up to date understanding of Google’s latest analytics property. To this end, we possess the knowledge and experience required to leverage the full power of GA4’s cutting edge features, granting our clients valuable insights and actionable analytics that help us drive real conversions and sales.
Superb Digital is an Bristol based SEO agency, we help businesses who manufacture and sell products online grow their revenue through a mix of data-led SEO, PPC, copywriting and digital PR strategies, utilising the most up to date tools and analytical tec...