We created a bespoke GA4 implementation style for Liberty London. This enabled Liberty London to adopt GA4 with total reporting parity vs Universal Analytics & put their best food forward on data collection while saving on on-site developmental costs.
Threepipe Reply is the long-term analytics partner of Liberty London and was tasked with conducting a full GA4 implementation based off the existing analytics.
Liberty London’s existing data structures were already comprehensive, so maintaining reporting parity was crucial. We also had to ensure that existing analytics were not affected, so both Universal Analytics & GA4 could run side by side.
The GA4 platform was just released out of Beta so documentation was limited but we knew it was crucial to begin the data collection process from day one.
Our bespoke implementation style was designed to reduce back end developer input by using JavasScript to translate existing data structures into the GA4 format. This also meant that a thorough page load speed testing process was crucial to the success of the project.
A successful implementation of GA4 consists of 3 key phases:.
Phase 1 – Discovery & Design
Phase 2 – Data Layer Validation
We adopted a “test and learn” GTM implementation of the new dataLayer to ensure that data coming into the test GA4 view was in line with our key objective of maintaining reporting parity between environments
Phase 3 – Production Testing & Final QA
This is all about the quality assurance of implementation. Our goal here is to ensure that the release of GA4 to the production Liberty side did not negatively impact the existing analytics of page loads speeds. We produced a speed testing framework to measure this and agreed release timings with all major Liberty stakeholders before publishing the final changes.
GA4 was deployed to the Liberty London site seamlessly with full reporting parity maintained across initial GA4 data collection. This early adoption of GA4 means Liberty London can:
- Begin the data collection process so reliable and optimised data is there for when it is needed
- Leverage machine learning for better predictions and insights
- Integrate seamlessly with BigQuery for custom analysis
- Move forward with confidence knowing that GA4 is designed to be flexible toward any upcoming environmental change around privacy (such as third-party cookie depreciation)
Image source: Getting Images – Eurasia Press