How to set up GA4’s custom insights and email alerts

What are custom insights?

Sometimes the data available in Google Analytics 4 (GA4) can seem a bit overwhelming especially if you have a complex site with lots of users coming and going every day. It can often seem difficult to quickly identify the most important trends or anomalies. Thankfully GA4’s Analytics Intelligence is here to help bring some much needed clarity. This is a group of features that uses modelling to help you understand your data. Insights is one such feature and it comes in two varieties, Automated Insights and Custom Insights.

GA4’s automated Insights use machine learning to detect trends and anomalies in your data and then presents them to you on the GA4 homepage and in the Insights dashboard.

Custom insights takes this further by enabling you to set your own conditions so that you can see the changes in specific parts of your data that are important to your business. Perhaps one of the best parts of this feature is that you can set up email alerts so that you and anyone else in your business can be notified when new custom insights are surfaced.

What is an anomaly?

Automated and custom insights can detect anomalies in the data. This is an exciting new feature for GA4 and is what makes insights so powerful, but you’re probably wondering how GA4 detects these anomalies.

GA4 uses Anomaly Detection to find abnormal data within a time series of data. This relies on machine learning predicting what the data should look like based on its observation of historic data. If the actual data appears a particularly long way off this prediction (data falling outside the credible interval) then it is counted as an anomaly.

A concept called principal component analysis (PCA) is also used to detect anomalies simultaneously over several metrics and dimension values at the same point in time. Dimensions and metrics are normalised by the number of users within multiple segments. If any segments show a significant deviation in the data and at least 0.05% of users on the property are affected then these segments are counted as anomalies. 

The direct benefit of Anomaly Detection for organisations is that GA4 can quickly find useful insights and potentially time critical anomalies in the data and present them to the user. This means the organisation can save time and resources that would have otherwise been spent on analysts painstakingly checking every possible combination of dimensions and metrics for these anomalies in the data.

Where to find your insights

When you open your GA4 property you will be taken to the home page. This contains some overview report cards on basic KPIs as well as a real-time summary.

If you scroll down to the bottom of this page you will find the ‘Insights and recommendations’ section. This is where you will see a selection of your automated insights.

To view and manage all of your insights, as well as create new insights, simply click on ‘View all Insights’ in the top right corner of the ‘Insights and recommendations’ section. This will then take you to the insights dashboard. If you click on an insight you will see more information appear to the right of the screen.

How to create your first custom insight

To create a custom insight you will first need to consider what kind of data is most important to your organisation and to any stakeholders who may access your GA4 property. Maybe there’s a particular type of conversion that is critical to your revenue stream or maybe your email marketing team wants to know about any drops in email acquisitions. However, if you are unsure and just want to get started straight away GA4 does also recommend some custom insights.

To start, you’ll need to click on the create button in the top right of the Insights dashboard. If you can’t see the create button or if it is greyed out you may not have a high enough access level on the GA4 property. You need to have either the analyst (or above) role to create a custom insight.

You’ll then be presented with the option to pick a suggested custom insight or start from scratch and define your own rules.

In this case we want to create a new custom insight from scratch so we’ll click ‘create new’ at the bottom of the page. Now we can set our conditions. The evaluation frequency will determine how often GA4 will check the conditions to determine if there has been a significant enough change in the data. You can choose hourly (web data only), daily, weekly or monthly.

Segment allows you to choose which groups of users should be included in the insight. This is done by setting conditions to include or exclude the relevant dimensions. Within the segment builder you can select up to 5 conditions. In our example we’re going to select ‘first user medium’ and set the dimension value as ‘email’.

The next step is to select the conditions required for your insight to trigger. You need to select the metric, which in our case will be ‘Total users.’ Next we’ll select the condition itself. This can be set to be an anomaly or any change by a certain count or percentage. If you’re using a percentage you’ll also need to set a comparison period to be either yesterday, same day last week or same day last year.

You’ll then need to name your insight. You’ll need to make sure this name can be understood by anyone who’s receiving the notification.

The final step is to select the email addresses you wish the notifications to be sent to. This can be anyone who has access to your GA4 property. In our scenario, for example, we would want everyone in the email marketing team to be notified when the insight is triggered. To do this we’ll simply enter each of their email addresses separated by commas.

Once you’ve entered all your email addresses you’ll need to scroll back up to the top of the page and click ‘Create’ in the top right corner to finish building your insight.

How to manage your insights

Managing your insights is easy. For automated insights you can leave feedback to help GA4 show the kinds of insights that would be most useful for you. This is done by hovering over the top right corner of the insight card and clicking either thumbs up or thumbs down. GA4 will then start tailoring your insights to be more relevant to your needs.

To manage your custom insights you’ll need to scroll to the bottom of the home page and click on ‘View all Insights’ in the top right corner of the ‘Insights and recommendations’ section. On the insights dashboard click on ‘Manage’ in the top right of the page.

You’ll then see a list of your custom insights. Click the 3 dots on the right of your custom insight. From here you can choose to delete or edit your insight.

Congrats!

If you’ve been following along, congratulations you’ve made your first custom insight in GA4. Your team will thank you when they get timely notifications on major changes in their data. In our scenario our email marketing team can breathe a sigh of relief as they’ll now get notified if their campaigns suddenly start underperforming!

If you’ve found custom insights to be particularly useful for your team we’d love to hear your tips on how to get the most out of this amazing feature.

If you need help getting started on your custom insights or just figuring out what you want to track in the first place, get in touch today.