In the wake of the General Data Protection Regulation (GDPR) and other privacy regulations, third-party cookies are becoming less reliable to measure key metrics like website traffic, conversions, and customer lifetime value. As a result, businesses are increasingly turning to behavioural modelling to track and measure user activity.
Behavioural modelling is a technique that uses data about user behaviour to create a model of their preferences and interests. This model can then be used to predict future behaviour, such as which products or services a user is likely to be interested in.
Tech marketers have incomplete data sets because most users don't consent to tracking.
GA vs GA4
Google Analytics (GA) is a powerful tool for tracking website traffic and measuring marketing campaigns. However, it has a major limitation: it can only track users who have consented to tracking. GA4's behavioural modelling for Consent Mode aims to fill these gaps.
GA4's behavioural modelling uses machine learning to estimate the behaviour of users who have not consented to tracking. The goal is to create a model that can accurately predict the behaviour of users, even if they have not given their consent.
GA4's behavioural modelling is still under development, but it has the potential to be a powerful tool for tech marketers. The accuracy of GA4's behavioural modelling depends on the amount of data that is available. The more data that is available, the more accurate the model will be.
What is Consent Mode?
Consent Mode is a new feature in GA4 that allows website owners to collect data from users who have not consented to tracking. This is done by asking users for their consent before collecting any data. Consent Mode is important because it helps to protect the privacy of users.
There are a number of different ways to collect data for behavioural modelling:
First-Party Cookies
One common approach is to use first-party cookies, which are cookies that are set by the website itself. First-party cookies can be used to track user activity on a website, such as which pages they visit, how long they spend on each page, and which links they click on.
Device Fingerprinting
Another approach is to use device fingerprinting, which is a technique that uses unique identifiers to track devices across different websites. Device fingerprinting can be used to track user activity even if they have cleared their cookies.
Once data has been collected, it can be used to create a behavioural model. There are a number of different algorithms that can be used for this, such as decision trees, random forests, and neural networks.
The behavioural model can then be used to predict future behaviour. This can be used to target ads, personalise content, and improve customer experience.
How to estimate data about user activity
Behavioural modelling is a way to estimate data about user activity, even when some data is missing. This can be used to answer questions about user behaviour, such as:
- How many active users do I have?
- How many new users did I acquire from my last campaign?
- What is the path that users take from landing on my website to making a purchase?
- Where are my site visitors from?
- How do mobile and web users behave differently?
Such insights help businesses improve their understanding of their customers and make better decisions about marketing, product development, and customer service.
Key Metrics That Can Be Measured
Here are some of the key metrics that can be measured using behavioural modelling:
- Website traffic: Number of visitors to a website.
- Conversion rate: Percentage of visitors who take a desired action, such as making a purchase or signing up for a newsletter.
- Customer lifetime value: Total amount of money that a customer is expected to spend with a business over their lifetime.
- Return on investment: Amount of money that a business makes from a marketing campaign.
Google’s behavioural modelling approach
Check for accuracy and communicate changes: Google uses holdback validation to maintain the accuracy of its models. This means that a portion of the observed user data is held back from model training and used to test the accuracy of the models. Google will also communicate any changes to the modelling approach that might have a large impact on your data.
Maintain rigorous reporting: Behavioural modelling is only included in reports when there is high confidence in the quality of the model. For example, if there is not enough consented traffic to inform the model, then events triggered by users who decline consent will not be reported. This helps to ensure the accuracy of the data.
Customise for your business: Google's modelling algorithm is customised for each business to reflect its unique customer behaviour. This helps to ensure that the modelled data is as accurate as possible for your specific needs.
Prerequisites for behavioural modelling
Consent mode must be enabled: Consent mode must be enabled for all pages of your sites and/or all app screens of your apps. This ensures that Analytics only collects data from users who have given their consent.
Tags must be loaded before the consent dialog appears: This ensures that Google tags are loaded in all cases, not only if the user consents. This is important for behavioural modelling, as it allows Analytics to collect data from users who decline consent.
The property must collect at least 1,000 events per day with analytics_storage='denied' for at least 7 days: This ensures that there is enough data to train the behavioural model.
The property must have at least 1,000 daily users sending events with analytics_storage='granted' for at least 7 of the previous 28 days: This ensures that there is enough data to customise the behavioural model for your business.
Behavioural modelling is a powerful tool that can help businesses to better understand their customers and improve their marketing campaigns. However, it is important to note that behavioural modelling is not without its challenges. One challenge is that behavioural models can be biassed, especially if they are not trained on a representative sample of the population. Another challenge is that behavioural models can be used to track users without their knowledge or consent.
Despite these challenges, behavioural modelling is a valuable tool that can be used to improve the performance of marketing campaigns and provide a better customer experience.