Google Analytics (GA) is a web analytics service that tracks and reports website traffic from users. It’s used across the world, and there are significant changes coming to GA next year.
This will impact everyone that works with GA. In this post, we’re going to explain how we’re approaching that change, provide details of how to learn more and how we can help people across government working with GA.
GA and GDS
GA is the Government Digital Service’s (GDS) primary source of performance data, which we have been building on over many years. We only collect data from users who give us their permission. The data is invaluable for understanding what is working, where the pain points are and how we can improve our platforms.
GOV.UK currently uses a version of GA called Universal Analytics (UA). However, Google is rolling out their new iteration called Google Analytics 4 (GA4).
What moving to GA4 means
We have to move to GA4 as data will stop being sent to UA in 2023 and the product will be shut down soon after. Google announced this back in March 2022, with their plans for ‘sunsetting UA’, which will happen on 1 October 2023 for paying customers (and on 1 July for non-paying).
The migration will have a big impact on how we measure performance data because:
- no tracking can be transferred, instead new code must be implemented
- every report and dashboard using GA as a data source will need to be rebuilt
- GA4 will be a new interface with new ways of working
Let’s dive into these 3 implications a bit further.
Firstly, the lack of “lift and shift”. GDS is looking into how to build an effective, efficient and intuitive tracking schema - which defines the structure of the data that you collect. More on this below…
Secondly, rebuilding datasets. Organisations will need to review dashboards and datasets, deciding if they should be rebuilt, by whom and who has long-term responsibility for the reports.
Thirdly, new software means learning new ways of working and training in best practices. As the software is still being developed, this will likely mean that the volume and scale of our reporting is likely to be different to what we’ve been tracking so far, so we’ll need to understand and control for these.
How we’re preparing for the GA4 migration
Due to the size of this piece of work, we began planning in September 2021 with support from Merkle (our Google products reseller) - initially working on the measurement design of our tracking schema. This is what we want to track and the best way in which to do that.
We began trying out the schema in our internal testing environment in Q1 this year. This was a low-risk way of seeing how it would work, without testing on actual users.
Doing it this way first is key for identifying any parts of the project we may have missed, the best way of working through the analytics, as well as aiding the writing of the Data Protection Impact Assessment (DPIA) on GA4.
What we’ll be doing next
Our next focus is around testing in Integration - the testing environment for GOV.UK. This is not only for quality assurance but also to check GA4’s value and usefulness for all disciplines within and outside of GDS.
We want to have this initial testing completed and our privacy notice updated this autumn, when we plan to put it live into our production environment on our "www.gov.uk" domain.
This deadline allows us one year of dual running of UA and GA4. During those 12 months we will have completed the following:
- comparison of the 2 data sources to understand how they are different
- further expansion of the tracking schema
- education and training on the new product and schema
- migration of reports and dashboards
- sharing the schema with other government departments
- rebuilding of the accounts, properties and views in a new, efficient way
- measuring how well it performs for users
- instilling best practices
We’re here to help
One of the aims of this project is to be able to share our tracking schema and our DPIA, in time, with other government departments.
This is an efficiency saving for government, and it’s also a major step in standardising analytics data tracking across the digital government landscape.
We will be blogging regularly about our work, and the best way to stay in touch if you work in government is via the Performance Analyst Basecamp where we will be posting regularly (information on how to join is on the Performance analysis community page).