Reviewing Channel Attribution & Campaign Performance in GA4
Google’s move to GA4 from Universal Analytics changes its attribution model, and you will undoubtedly see a difference in how revenue is attributed across your digital marketing channels based on the new definition of a ‘session.’ After migrating to GA4, you will see a shift in your marketing results even without changing anything to your marketing spend or strategies. But where do you go from there, and how do you properly analyze your data to make the best decisions for marketing your brand?
What Should I Do with My Universal Analytics?
Before your Universal Analytics data becomes inaccessible, we recommend keeping everything active and extracting as much as possible – this includes marketing acquisition data, behavior data, and event data. You can also pull audience data, but this does not change significantly over time. Listrak’s VP of Strategy and Analytics, Jake Pasini, walks through his specific recommended steps within Custom Reporting.
You can export your Universal Analytics data by week, month, or however you need it organized, but make sure you’re always getting a non-sampled dataset for a full, accurate view. Depending on the cardinality of your data (how many permutations of source/medium, campaign, and other key variables), you may want to pull the data by week or month rather than day to make the row count more manageable.
You’ll always want to pull raw data as compared to calculated metrics and then do the aggregation yourself wherever you’re storing these results: think unique page views and bounces as metrics to then calculate bounce rate, as compared to exporting the bounce rate percentage.
What Are the Main KPIs I Should Focus On?
The three main data points you should strive to export from Universal Analytics are sessions, transactions, and revenue. These allow you to then calculate and store conversion rate, AOV, and other KPIs. For marketing data, you’ll want source/medium, campaign, and potentially ad content or keyword if valuable to your business. In GA4, you’ll want to adjust your attribution setting to last click for a cleaner comparison to the data you exported from Universal Analytics and change your data retention setting to 14 months from the default 2 months. This opens a more extended period to collect results in your custom reports. Note: if you do not change this setting then ANY custom report will only be able to go back 60 days from today!
So, all the cool functionality of ‘pull any data point you want’ or ‘may any type of viz’ will only be for the last 60 days. (Stop reading this and go change it now – please!)
What Is Defined Differently in GA4?
Sessions are defined differently in GA4. New sessions are no longer created by the clock hitting midnight or UTM parameters changing, but rather when a user enters your site or app, and a session isn’t already active. A session times out automatically after 30 minutes of inactivity, and there’s no maximum time or limit a session can last. Most importantly, this essentially changes a session from ‘last click’ to ‘first click.’ You can no longer look at sessions as a proxy for clicks, and paid media will get a boost as it is a major first-click channel.
Bounce rate is also different in GA4, and traditional ‘Bounces’ found in Universal Analytics no longer exist. Instead, retailers can look to engaged sessions, engagement rate, and bounce rate to analyze their user engagement. An ‘engaged session,’, is when a user has spent 10+ seconds on a page, has completed an event (i.e., purchased), or visited 2+ pages.
GA4’s User Acquisition report has a First User model that redefines how channels get credit for acquisition. It will look for the first time a user is ever identified on your website and attribute all their activity thereafter. It’s not the most accurate or reliable way of looking at channel performance and something Listrak does not recommend using. We have seen numerous online-only businesses have Email and SMS attribution still significant portions of revenue, but it is almost impossible for the first way someone comes to the website to be those channels as you must come to the website to sign up. It was honestly an impossible task for Google and isn’t very usable unless you are starting a brand-new website.
What Deltas Can I Expect to See When Comparing My Campaigns?
In our comparison exercises with clients so far, we’ve seen deltas all over the map with the different definitions for sessions and the construct of the engaged user. But in general, we’ve witnessed Email and SMS down as much as 30% lower up to 15% higher – It’s the wild west with GA4’s new systems! And remember, this is without changing your marketing spend or strategies.
Listrak is Here to Support You
Listrak is a trusted partner in your migration to GA4 and understands it will significantly change how you analyze your digital marketing performance and overall business health. It’s essential to thoroughly understand the differences in your campaign results before making any drastic changes and set new goals and expectations with your leadership teams based on where you want to focus. Be sure to check out our GA4 Migration Resource Hub for more information and tips.