Skip to main content

Now, with just a few days left before the Universal Analytics API deprecation, I am wondering how everyone dealt with the GA4 transition in their reporting. Did you keep using bounce rate, or did you switch to looking at engagement rate and engaged sessions?

This has been a tricky line to navigate while re-creating our old Universal Analytics templates with GA4. Do we keep the metrics just as they were in Universal Analytics? Do we use a mix of both the old and the new metrics? Or do we embrace the change and use the new GA4 metrics exclusively?

I'm curious to hear your thoughts.

Maybe I’m in the minority here, but I was glad that bounce rate disappeared, and advised clients to steer clear of that in the future, as there was no word on it ever returning down the line. The main problem we saw was that some clients had super low bounce rates (driving to blogs, product pages, etc.), and some clients had super high bounce rates (driving from landing page to third party sign-up platforms), which made benchmarking and comparative analysis over time next to impossible. Some were valuing too heavily as an indicator for success, when it made more sense to shift focus onto how people were using their website to convert, not generally how people were using their website.

What we did was implement custom metrics that focused on their key business goals, then aggregated those key actions (before conversions became key actions) into a custom engagement rate. For example, if they wanted to know how many people were fully scrolling their page (scroll depth = 90%, tracked by default as ‘Scrolled Users’), how many people clicked a ‘Learn More’ button (custom click action tracked through Tag Manager), and how many people got to a download thank you page (custom action tracked through Tag Manager), we would build those out, then in Looker Studio, use a simple math function to sum those into ‘engagements’, and then divide that by total sessions for an ‘engagement rate’. Clients like this because it’s transparent, makes sense for their business, and cuts down on the confusion when relaying message up the stakeholder chain.

The update forced a lot of clients that weren’t looking at custom events to forgo their old UA setup and lean fully into GA4, and in truth, a lot of their data from 2020-2022 was skewed from the pandemic anyway, forcing them to re-establish benchmarks and goals with how their businesses evolved. Some of these teams also pay a lot for their website vendors and can’t afford to make big UI/UX changes, so having these custom signals allows them to open that dialogue, create a business case for making the changes, and optimizing when they can to help support the business as it grows/evolves.


Wow, thank you for sharing!  It’s interesting to hear that you saw your clients actually re-evaluate their benchmarks and goals to adapt to the changes. 

It’s also great that you were able to leverage website events to create a custom engagement metric that focuses on how people are using the website to convert, rather than just how they use the website. This is a very good tip for those building their own reports ✍️.


Reply