Storing & Blending Google Analytics Data

On July 1, 2024, Google will stop storing data from Universal Analytics (UA). This shift signals the final migration from UA to Google Analytics 4 (GA4) and the end of all UA data. It means that all metrics, including pageviews, engagement rate, and conversions, will be lost in the sunset—unless you export and store your data in another location.

Additionally, Google has updated its data retention settings in GA4 to include expiration caps. No longer can you set your data to never expire. Instead, you must choose an expiry time between two and fourteen months in which your GA4 historical data will reset. At the end of the period, you’re back to square one.

This loss of historical data creates a huge problem for companies who depend on year-over-year comparisons and other crucial metrics to inform their marketing decisions. Fortunately, Gemini’s clients are safe. Not only have our data engineers discovered how to safely back up all historical GA4 and UA data, they know how to blend it with incoming data from multiple sources. This allows us to create seamless reporting that provides our clients with the most comprehensive snapshot of their marketing efforts. Let’s look closer at this process.

How to Export and Blend UA and GA4 Data: 3 Phases

Phase One: Migration

In order to move data from one location to another, you first need to identify which datasets you’re transferring and then prepare them for their journey. At Gemini, we saw this process as an opportunity to warehouse not just Google Analytics data for all of our clients, but to direct all of our tracked data into one location. This meant identifying and adding data from CallRail and other applications to the mix. Once all of our data was funneling together, we could then revamp our client reporting process to provide a more holistic view of each client’s marketing program.

Our Migration Process

We turned to Analytics Canvas as our migration application. It has a built-in interface that supports Google Analytics and allows you to manage large volumes of data from multiple accounts. It’s perfect for a marketing agency with a large roster of clients, each with their own datasets.

We set up our Analytics Canvas account to include tracking for paid and organic traffic and integrated the platform with our other tracking applications. As of July 1, 2023, Gemini was backing up all marketing data for our clients. We also designed the new system to evolve as we added data from additional sources and incorporated new metrics from GA4.

Phase Two: Data Blending

Having all our data safely stored in one location was a major victory. But, we needed to organize and blend that data before it could provide the deep marketing insights our specialists require to make their program decisions. Without proper blending, the data would be inaccurate or have gaps. Our blending process involved mapping, filtering, and testing the data before it could be compiled for reporting.

Data Mapping

Mapping the path of each dataset was a long and tedious task. It required us to meticulously chart each data source from every client. Our visual map consisted of connected blocks that represented data, filters, or blending mechanisms. The result was a document that detailed each dataset’s journey from its origin to final blend.

Filtering

Filtering improves the accuracy and efficiency of the data by removing errors and undesirable elements so that only the common dimensions remain. Each data stream needed to be filtered before it could be blended with other streams. For this stage, we used the Visual SQL feature in Analytics Canvas. After the data traveled through Visual SQL, we performed a final filtering as we imported the data into Looker Studio for reporting.

Testing

Once the data was blended, we tested the results for accuracy before utilizing it in reporting. Since there are numerous ways to track the same data, we tested each. After a month, we compared the results of these tests to existing data and determined which tracking methods were the most accurate.

Phase Three: Reporting

With all of our client’s data properly migrated, blended, and tested, it was fairly straightforward to customize and automate reporting for each. For this process, we utilized Looker Studio. As an extension of Analytics Canvas, Looker Studio provides a dedicated interface with GA4. It allows users to analyze metrics and compile the results visually in easy-to-digest reports.

Now, every day, we automatically backup, blend, and filter UA and GA4 data and export it to Looker Studio. Read Gemini’s article on marketing reporting to learn more about this process and the value of our revamped reporting.

Are Your GA4 Datastreams Safe?

If you’re one of Gemini’s clients, your data is safe. If you’re not, your datastreams are at risk of disappearing every time analytics resets. Not only can we assist your company with preserving and blending your valuable historical data, we can provide you with the most comprehensive marketing reporting available. Contact Gemini today to learn more.