The Difference Between GA4 and Universal Analytics

Google Analytics 4 (GA4) is designed to collect data differently to universal analytics (UA / GA3). The change in collection method means that behaviour is measured using an event-model and users can be more easily recognised even when coming from different devices and across different touch points (say an app and a website). There are also some metrics being dropped with GA4 putting more focus on engagement and revenue rather than general behaviour. As a user of Google Analytics, these are what the changes will mean for you:


  1. User and session reporting will be more accurate (so be prepared to explain to stakeholders why these values may be lower than in UA)
  2. Engagement metrics show what’s really working – no more inflated bounce rates due to people clicking on order tracking pages.
  3. In theory attribution modelling will be more accurate with machine learning being used to augment actual data
  4. Simplified dashboards deliver top-line insights quicker across the customer lifecycle
  5. The data explorer makes it easy to build custom reports and these can be saved and shared so everyone is looking at the same data
  6. Some performance data is still accessible and machine modelling fills in the gaps as consumers become more privacy conscious.


Google has released a detailed document explaining the differences in data collection in this article. Rather than rehash their information, we’ve decided to focus on how businesses can use the launch of GA4 to start thinking about improving their performance reporting.


Where did you come from, where did you go? It’s not just Rednex who wants to know this. Most marketeers have become very comfortable with the Universal Analytics “last-non-direct-click” default model of attribution. This means that in most scenarios the visitors and conversions are assigned to the last source they clicked on, unless it is a direct click. The only time direct is logged as a source is when there are no other preceding sources. In most cases this gives a good indication of channel performance at a top-level, however, if you are a brand where users are likely to make multiple purchases or visits in a short space of time for different reasons then this can be misleading. It also doesn’t show purchase hesitancy.  

GA4 shakes things up a bit by showing a number of default views of acquisition data with session and user parameters. The user acquisition card in the acquisition dashboard will show the originating source for users who were active in the given time period. This means that if the time period is set to April and a user came in via email, but the user first visited in January from organic search, the credit will be for organic search not email. The traffic acquisition report however will show the source of the sessions for the users in the given time period. Those in charge of channel performance will need to be aware of these differences and be intentional about how to report on performance going forward. This is especially important as each channel will also provide their own reports using different collection methods and views. Having clarity about what your business considers attribution and what metrics to use to model this will make it easier to judge success in a consistent way.

Actionable Insights

Are you still capturing data from multiple sources to send out daily KPI emails? With Google’s minimalist approach to reports, it’s time to become inspired to be intelligent about how we get to the right level of insight needed to make decisions. GA4 now has custom insights ( which enable marketeers and traders to set up email alerts. For example, to be notified when traffic drops below a certain threshold. It is much better to focus communications on “things that need action” rather than a daily update that becomes routine and easily ignored. GA4 also now provides some predictive capabilities that can help identify purchase and churn probability and predicted revenue. This can be used to get ahead of potential issues rather than reacting once they have happened.  

Dashboards can still be set up in data studio or GA4 for those that want somewhere to look at top line metrics, but now focus can be on future actions rather than retrospective data.

Transactional Analysis

It’s time to move away from Google for transactional analysis unless there is a attribution dimension to the analysis. This means it is fine to see roughly what channels are contributing to sales and if there are any obvious trends in purchases by source (E.g. Instagram users are buying dresses but tiktok users are buying shoes). However, detailed basket analysis, accurate sales data and customer value data should be coming from your source of truth for sales data. This might be your ecommerce engine, your OMS or your ERP. This is because transactional analysis should include all orders and Google Analytics will not have this. Some orders will not be captured in GA4 due to consumers privacy choices and potential technology failures. Transactional data can easily be fed into Big Query to augment Data Studio reports if there are no other ways to visualise this data within your business.  

What about as a merchandiser? If you are still using segments and ecommerce reports to merchandise category pages, GA4 is going to make it harder to find the data needed. Especially as GA4 exploration tool doesn’t provided calculated metrics (e.g conversion rates). In reality merchandising should be done with specialist search-and-merch tooling, however, we appreciate not all sites have this available. Merchandising is no longer about viewing each page as a stand alone entity but as a stop-off in a journey driven by user needs and wants. As such, the improved path exploration tool ( will be useful. Remember to make sure site search is being tracked if you have it. GA4 will track 5 query parameters by default (q, s, search, query and keyword) so you may need to configure a custom parameter if the site search doesn’t have one of these values after the ? on a search result URL.

Still needing to migrate to GA4? Or need help improving your teams confidence with data to make good decisions?

Contact one of our data team for a chat about how we can help