With the switch from Universal Analytics to Google Analytics 4 fast approaching, we’ve created this checklist to help you prepare.
It’s in order of priority and has taken into consideration that some tasks are dependent on others. We’ve differentiated between the tasks that are accessible and those that will require specialist support, and we advise you to lean on your internal data team or external agency partner for help.
Download your copy of the checklist by clicking here or view it below. 👈
- Ensure website tagging is GA4-ready (via Google Tag or GTM) [specialist]
- In admin settings, check that your data stream is sending information to the GA4 property [accessible, specialist support may be required if it’s not working]
- Set up relevant Google product links [accessible]
- Ads and/or Search Ads 360/Display & Video 360
- *Search Console
- *Merchant Center
- *BigQuery
- *Play
- Configure your attribution settings [accessible]
- Attribution model (use Cross-Channel Data-Driven)
- Look-back window (use maximum time, unless you have a specific reason to narrow it)
- *Enable key GA4 features, if you’re happy to do so [accessible]
- Google Signals for audience exports and cross-platform reporting
- Enhanced Event Measurement to automatically collect user journey data
- *Check that the events section is populated with everything you want to track [checking is accessible, repopulating/customising likely needs specialist support]
- Mark relevant events as conversions
- Check that your purchase event data aligns with your expectations (especially in revenue terms)
- Import your GA4 conversions to Google Ads as secondary conversions to quality check their data. Aim to use these as your primary conversion actions before July 1st. [accessible]
- (There might be some people using the Google Ads tag with Enhanced Conversions who want to continue to use that, which is a completely valid choice)
- Build audiences ASAP to start collecting data ahead of July 1st, as they are not retroactive. If you build them after July 1st, they will not be immediately usable. (Note: max 100 audiences per GA4 property. However, you can always archive old audiences to build new ones) [accessible]
- Remarketing audiences (e.g.) High engagement without purchase / cart abandonment
- Exclusion audiences (e.g.) Recent purchasers
- Reporting audiences (to use with comparisons) (e.g.) Specific sequential on-site actions, to monitor site flow
- *Access the report library to tailor your GA4 interface [accessible]
- Explore the default report templates. Some templates will be created based on your linked accounts (like Search Console)
- Create custom reports that answer specific questions, or offer insights relevant to broader business objectives. Keep these simple and succinct.
- Create collections in the report library – this enables you to group sets of reports together, and they will appear under subtitles in the navigation bar in a similar style to the Universal Analytics UI
- *Build Exploration templates that will be useful for deeper analyses in the future (or just for the sake of limit testing GA4!) [accessible]
- We highly recommend that eCommerce businesses build a simple Exploration to show revenue alongside the Default Channel Group and/or Source/Medium dimensions. Don’t use the variations of these dimensions that have prefixes like “Session” or “User.”
- This exploration shows you revenue through the lens of GA4’s Data-Driven Attribution, which gives a better idea of multi-touch journeys than last click.
- When using UA, you might have developed a habit of checking the channel group or source/medium reports to review grouped channel performance. In GA4, the base “Session Acquisition” report still uses last non-direct click data because it is session-scoped rather than event-scoped.
- Make sure you’re taking full advantage of GA4’s attribution model by instead leaning on our recommended Exploration to review channel performance. This uses event-scoped data, which is the beating heart of GA4.
- *If you understand and accept the privacy implications, set up Google’s Consent Mode, which will allow GA4 to start modelling data lost to users that decline cookies. This is really valuable information that can impact decision-making [specialist]
- Read up on how the data modelling works – it’s a good methodology, but it is still a model. Some stakeholders may require an explanation or justification
- You can compare modelled data with your actual overall business data to establish a sense of the model’s accuracy
- Understand the different ways that GA4 and UA calculate and display information [accessible]
*these deliverables could wait until after the switch, but it’s best practice to have them in place now.
If you have any questions or require support ahead of the switch on the 1st of July, get in touch with our team. For more Google Analytics 4 resources, click here.
