Google logo on fence
PHOTO: Kai Wenzel | unsplash

Analysts these days chase down data from untold number of sources and depend on visualization tools to tame the data flood. Now Google has made its bid for Google Analytics 4 to be a top contender among convenient visualization tools by allowing data to be imported into the platform for the first time.

The feature represents a major shift for the tool. Since Google Analytics' early days it was possible to export data as a spreadsheet, CSV file or as a PDF. Other platforms have also introduced plugins and extension to export data through the Google Analytics API, opening the door for a wide variety of export applications, such as developer access through the API, plugins for Excel from SuperMetrics, or libraries within programming languages like R programming. Even Google itself came up with a connector for its own cloud platform, Google Data Studio.

The import feature introduces a whole new level of convenience, permitting analysts and marketers to see what data blends with Google Analytics data. When considering adding any new data solution to a workflow, data analysis team must bear in mind what type of visualizations are possible from the imported data. Examining the output of visualization can help a team decide which kinds of extensions they need. Importing data into Google Analytics allows analysts to decide on their visualization needs faster. 

How Imported Data Is Treated

In Google Analytics, data is joined between event data recorded from the analytics tag and your data. Users access the import under the analytics property in the admin section. The navigation in the Google Analytics interface is Admin > Property > Data Import. The data import feature can incorporate either spreadsheets or CSV files. The types of data can be user-related, such as a remarketing list based on a loyalty rating or CRM-produced data, or product-related, such as a merchandise list.

GA4

Users then select a source — a container for the imported data. The source can be new or a previously saved source. When selecting a new source, users can add a descriptive label, select a data type, and map the data in the source. Mapping involves highlighting what is a key dimension from the data so that the other dimensions align correctly.

The data types determine the two joining methods available called reporting query time and collection/processing time. A product data type is applied to the reporting query time method. The reporting query time method creates a join when an analytics report or query is initiated. The join is temporary, eliminated when the import file is deleted.

For the second data type, user attributes, the collection processing time selection is used. Data is joined according to the time period in which the analytics data is processed.

One thing to know with this method is the user cannot also then join to historical data, something to keep in mind when deciding how to approach an ad hoc analysis. Analysts will have to plan anticipated time periods to add data, like an upcoming sales season where data sourced outside of GA4 would potentially fit an analysis.

Related Article: How to Use Cohort Analysis in GA4

Limitations to Consider When Importing Data

There are limits to the amount of data that can be imported. The source size can only be up to one gigabyte. There is also a daily upload limit of 24 uploads, with a cumulative total storage of 10 gigabytes. This means that analysts will either have to manually import, or export Google Analytics data into another platform like Google Data Studio to handle data delivered as a stream.

Another deciding factor for relying on imported data is what kind of graphics demands your team receives on a regular basis. Some visualizations will work well within the solution you have — in this case, GA4 — while others may need a different venue to get the visual you want. For example, you may need some statistical metrics alongside your data, such as viewing boxplots to see if data contains outliers. In that case, you would want to import data into programs such as R programming or Python models.

In the case of Google Analytics, joining data does not replace a statistical analysis, yet it does still offer the convenience of seeing if that joined data can be quickly visualized or worth investigating later in an advanced a graph or a model, which can save some development time.

Overall, the Data Import feature is a necessary upgrade for data management within Google Analytics. With applications evolving way beyond websites, and the rise of privacy compliance, analytics must offer more convenient ways for users to better manage the data they import.