Google Analytics 4 (GA4) was first released in October 2020, and Google has added several updates this year already. My overall impression of GA4 is that it seems targeted toward enterprise and ecommerce sites, but may not be as useful for smaller businesses. At the time of writing, I don't see much merit in switching to GA4 already, and I think many digital marketers would agree.
GA4 has gotten a lot of negative reactions, mostly from digital marketers saying that it's difficult to learn and use, even for seasoned professionals who are usually quite tech savvy.
New Features in GA4
GA4 offers a lot of new features that were not previously available with Universal Analytics (UA). It's not just a new layout, but a completely new event-driven data model and interface, designed to get digital marketers and strategists thinking about how website and app engagement relates to the customer journey.
An increasing number of businesses are using multiple digital touchpoints to connect with their customers. For example, you might have both a corporate website and a mobile app. Cross-domain tracking was possible in UA, but it required more resources, and could be complicated and error-prone with coding or GTM processes.
You can now set up cross-domain tracking directly in GA4 without relying on Google Tag Manager, so that multiple domains are included within one property, resulting in less resources needed for both implementation and analysis.
UA was never quite useful for analyzing the online customer journey throughout your site. As a web analytics specialist, I often received requests to analyze the most common page paths that led to a conversion, which pages caused near-converting customers to drop out, etc. The analysis was possible in UA, but there were a lot of separate parts and it definitely wasn't a smooth process.
With GA4's funnel analysis feature, this problem should be resolved. You can see the customer journey for different user groups, including backwards analysis (starting from the conversion event and looking back at the previous visited pages), as well as what non-converting users did when faced with some error on your site or app.
GA4 has added many AI-based features, and anomaly detection is just one of them. If you also work in PPC, it's kind of similar to the anomaly detection feature of Google Ads. For example, if Google had predicted that your ecommerce site will have $10,000 profit today, but it only had $1,000 at the end of the day, that's an anomaly. This eliminates the initial need for you to check and determine if something was statistically significant, because Google will tell you for free. However, you aren't completely in a black box relying on Google's algorithm. In GA4, you can manually turn on anomaly detection for line graphs, set the sensitivity and learning period.
You can now connect GA4 with BigQuery for free, as well as export to BigQuery Sandbox. This allow you to go beyond GA4's reporting interface and use the advanced analytics features in BigQuery instead. For example, you can connect GA4 with third-party APIs, export custom event parameters and dimensions, and push data to visualization tools such as Google Data Studio, Tableau, and Power BI.
GA4 is still under active development. Some things are still missing, such as attribution, tool integrations, and detailed ecommerce features. These might get added in the upcoming months.
Migration to GA4
There won't be a forced migration to GA4 in the near future, and you can continue to use UA. Unfortunately, Google is not connecting the UA and GA4 codes. So it's recommended to create a GA4 account and accumulate historical data in it though, even if you don't plan to actively use it. That way, you'll be able to conduct web analytics whenever you do eventually switch to GA4.
For websites that already have UA, you can quickly make a GA4 property in the admin settings. On the other hand, for completely new website launches, you might be ok just starting with GA4 and don't bother with UA anymore. While you're waiting to accumulate data, you can play around with the Google Analytics demo account to see how things work.
You can create a new UA or GA4 property, or both, from the admin section in Google Analytics. Go to create a new property, then show advanced options.