Why do the numbers never match in GA4?
What are the types of data in Google Analytics 4 and why do they seem never to match?
Main Takeaways
GA4 Standard Reports
Standard reports are designed to show trends. They can be used to see the ebb and flow of the data for your website or app. They are not intended to be a source of absolute truth in GA4. This is by design in order to make the platform efficient.
GA4 Exploration Reports
Exploration reports are the tool used to take a deeper look in to the data when users wish to understand the trends they are seeing. The data in the lines of the table is not aggregated however, the totals may be aggregated. This means the totals may not match the sum of the lines. It is good practice to export the report as a spreadsheet and re-calculate the totals and interogate the data in your chosen spreadsheet.
GA4 Reporting API
Reporting API is used to automate the creation and delivery of reports similar to those available in the Exploration report section. They can pulled into other platforms directly or the data can be delivered in whichever way you wish.
GA4 BigQuery Connection
BigQuery allows the collection of complete data sets, and has a report creation function. BigQuery is the source of complete truth in GA4. Best practice is to
Ever wondered why the numbers never match in your GA4 reports? The answer may be more simple than you think.
When Google created Google Analytics 4 (GA4) to replace Universal Analytics (UA) they started from the ground up. The GA4 platform is a complete rebuild which includes many great additions however, one of the priorities was to reduce the server resources required to run the platform.
Among the solutions used to achieve this were the aggregation of data, sampling and limiting the number of rows in a report table.
Aggregated data is composed of individual user data consolidated at a broader level, making it impossible to link to any specific individual. This amalgamation (aggregation) occurs without including user IDs or event timestamps, rendering the data a collective overview of user activities. To illustrate, the Pages and Screens report provides the average engagement time for each page, yet it does not disclose the duration an individual user spent on a page.
In Google Analytics, data sampling might occur if the quantity of events utilized to generate a report, exploration, or request surpasses the quota limit assigned to your property. In such instances, Analytics employs a subset of the data and subsequently extrapolates to offer results that are directionally accurate and representative of your entire dataset.
The limiting of the number of rows in a report can lead to the phrase “other” being used to define less prevalent elements, for example, the lesser-used channel groups may be grouped together and described as “other”.
This is often thought by users to be an error in the setup or in GA4 itself, however, it is not. It is a fully thought-out and designed feature of GA4 aimed at increasing the efficiency of the platform. Having said that there are best practices to limit the amount of times you see “other” in this way.
GA4 has several methods to produce reports these are the standard GA4 reports, exploration reports, pulling data via the API and exporting to BigQuery, and producing reports from there.
Each method uses data that have specific characteristics for example:
As you can see from the graphic above BigQuery is the best way to record and report on your GA4 data.
Google Tag Manager Solutions
Enhanced Ecommerce Data Layer
Recommended Events Data Layer