2023-01-12 | Article | Insights
An accurate, reliable, and relevant data foundation is necessary in order to make data-driven decisions. The ability to obtain and maintain top quality data in GA4 will enable one to make confident business decisions and unlock the power of GA4 for answering fundamental business questions.
The challenge with any kind of analytics data is that still today, it is not prioritized high enough within certain companies or organizations. This can have multiple reasons, one of them being the lack of accurate data collected and therefore its lost credibility. On top of this, GA4 in specific is based on an entirely new data model, which demands new know-how and many organizations currently struggle in their transition from Universal Analytics to GA4, as its implementation remains complex. As a result, analytics projects tend to stagnate quicker than they should.
No implementation is perfect. The goal is to minimize the errors and maintain accuracy. The process to do this depends on where one stands with their GA4 implementation. If GA4 has already been implemented, the data quality can be assessed by auditing the current GA4 setup. If one hasn’t started with GA4 yet, then it is easier to start data collection from a blank slate rather than dealing with the consequences from a faulty implementation. One way or the other, once data is being collected, its quality needs to be monitored continuously just like with any kind of system.
In this article, we show which steps need to be checked in order to achieve and maintain high quality data in GA4:
✓ Audit GA4 setup
Audit the GA4 data with the guidance of a health check report. The report summarizes what works properly and what is missing, providing an overall GA4 quality score and giving focus on what to work on.
✓ Define business goals and KPIs
Have a clear understanding on what one wants to track. Only collect data that is relevant.
✓ Implementation specifications
Technical specifications for GA4 tracking in documented form is the basis for developers to implement tracking. The specifications need to be free of errors as they form the basis of data collection.
✓ GA4 / GTM configuration
GA4 properties and streams need to be correctly set up. GTM or any other Tag Management System needs to be configured correctly so that it can receive and forward data to GA4.
✓ Testing
After the implementation and configuration, tracking needs to be tested on a testing environment first (e.g. staging), followed by testing in the live environment. All possible scenarios and user journeys need to be considered and tested thoroughly with the right tools and technical knowledge as code and logs are being checked. Small mistakes such as typos in values of parameters (e.g. price value wrongly formatted in ecommerce) can have a big impact on data quality. Nevertheless, if the mistakes are discovered in a testing environment or before the release to the live environment, mistakes can still be fixed with little effort. The costs of careless testing are considerably higher when data collection already takes place or, when business decisions have already been made.
✓ Check reports
Check the reports in the GA4 interface to check whether the data makes sense and is what is expected. Even better, connect the GA4 data streams to a data visualization tool such as Looker Studio to create more advanced analysis and visually appealing charts that help one understand the data better.
✓ Monitor data quality
Set up custom insights in the GA4 interface to receive notifications to email when something unexpected happened (e.g. traffic drop, drastic change in traffic source, higher or lower conversions etc.). Custom insights give the opportunity to quickly identify anomalies in the data. Furthermore, continuously monitor the GA4 data with the help of a quality dashboard in e.g. Looker Studio to keep track of important GA4 quality metrics and dimensions.
✓ Campaign tracking
Are campaigns being tracked? In GA4, there is currently no customization of the channel grouping possible. That means that custom namings for medium will be bucketed into “other”. Hence, one should stick to the Google predefined namings for medium for the time being. Other UTM parameters such as campaign and source are still customizable (more about Channel Grouping in GA4 here). Campaign tracking is straightforward, but needs to be done methodically and ideally with the help of an automated campaign builder to avoid error.
Follow the above steps to stay on top of GA4 data quality but keep in mind that this is not a one-time job. Just as much as the web changes, so does the technical landscape, website content, marketing campaigns and so on (e.g. think of ecommerce websites in particular). If data is not reevaluated again, it is very likely to deteriorate over time.
For these reasons and more, GA4 data needs to be audited regularly, and how often that is, depends on the size of one's tracking and the organization. We recommend checking the GA4 setup at least once a year for smaller tracking setups and several times a year for larger setups. Audits also make sense to be done at specific times, e.g. when stakeholders change or technical landscape develops. Ultimately, there is not a size fits all approach.