For those who had an opportunity to learn earlier guides we’ve revealed on Google Analytics 4 (GA4), you most likely know that it’s not a plug-and-play analytics device like Common Analytics was.
There’s a whole lot of data one wants to soak up so as to have the ability to arrange GA4 correctly, and time is ticking.
With GA4 being a extra complicated device, it’s simple to make errors that may hinder the accuracy and reliability of the information collected.
On this article, we’ll discover 5 frequent Google Analytics 4 errors that may simply occur and supply sensible tricks to keep away from them.
1. Not Setting Information Retention Interval
GA4 comes with a two-month knowledge retention interval by default, and you’ve got the choice to set it to 14 months. The retention interval is utilized to customized studies in explorations, whereas knowledge in the usual studies by no means expire.
As soon as the retention interval has handed, knowledge will probably be robotically deleted – which suggests should you don’t change that setting as you arrange GA4, you will be unable to run YoY customized studies and can lose worthwhile historic knowledge.
As a way to change the retention interval, navigate to Information Settings > Date Retention, and within the dropdown, select 14 months.
Additionally, you will discover a checkbox studying “Reset person knowledge on new exercise,” which suggests the 14-month knowledge retention interval is counted from the second of the person’s final go to onwards.
In different phrases, every time a person engages in a brand new exercise, their knowledge retention interval will get prolonged for an additional 14 months.
Truthfully, I can’t consider a use case whenever you would select to show that choice off, so I maintain it switched on.
2. Dimensions With Excessive Cardinality
Excessive-cardinality dimensions are dimensions that include greater than 500 distinctive values inside a single day. This will current challenges and limitations in knowledge evaluation inside GA4.
Cardinality in GA4 can negatively have an effect on knowledge accuracy and reliability.
For instance, whenever you observe the precise phrase rely as a customized dimension on each article web page, chances are you’ll find yourself having excessive cardinality if in case you have hundreds of articles as a result of the phrase rely may be totally different for each article.
How To Repair Excessive Cardinality
To mitigate the influence of excessive cardinality in GA4, contemplate making a bucket of values.
With the instance above of phrase rely customized dimension, it actually doesn’t matter that a lot whether or not the article can have 500 or 501 phrases. You’ll be able to bucket values into ranges like:
And as an alternative of pushing too many distinct values, you’ll have solely 5 totally different dimensions.
Additionally, as a greatest apply, at all times outline customized dimensions properly.
Make sure that customized dimensions align together with your evaluation goals and contemplate their potential influence on knowledge accuracy and useful resource consumption.
3. Not Linking To BigQuery Account
Linking to BigQuery was out there in Common Analytics 360 however not within the free model. With GA4 now, all customers have entry to that premium function.
Since it’s exporting knowledge to BigQuery from the second you join, it’s essential to set it up at first to be able to have as a lot historic knowledge as potential.
BigQuery has a giant benefit over GA4 customized studies as knowledge is rarely sampled, whereas in customized studies, knowledge will probably be sampled if there are greater than 10M occasions within the exploration report.
As a way to hyperlink GA4 to BigQuery, navigate to BigQuery Hyperlinks in your GA4 settings.
As a way to full linking to BigQuery, you would want to create a BigQuery undertaking which would require you to enter your billing data.
It’s freemium, and 10 GB is free every month; it’s going to cost you $0.02 per GB should you exceed that quantity.
4. Failing To Set Up Customized Audiences
GA4 has highly effective audience-building capabilities you’ll be able to learn extra about in our information on the best way to create segments and audiences.
With GA4 audiences, you’ll be able to analyze particular knowledge segments, empowering you to derive worthwhile insights. As an illustration, you’ll be able to create goal audiences corresponding to engaged customers, subscribed customers, or customers who made a purchase order up to now 30 days.
It’s advisable to create audiences to your ICP and mark it as a conversion.
Since viewers dates are usually not retroactive, you will need to outline your goal audiences at first of setup to be able to collect historic knowledge.
5. Utilizing Auto Migration From Common Analytics
GA4 is a very totally different beast in comparison with UA, with a special knowledge mannequin.
Despite the fact that it provides the choice to robotically acquire Common Analytics occasions, it’s higher to not use that, as it’s a likelihood to rethink your analytics and design your occasion assortment structure anew for higher analytics.
6. Not Excluding Undesirable Referrals
Usually ecommerce web sites have third-party fee processors that are hosted below totally different domains – and when redirecting them again to the web site after the person completes a checkout, GA will detect it as a brand new session as a result of the referral is totally different.
As a way to keep away from that and never distort your conversions knowledge, you must exclude such domains from referrals so GA doesn’t provoke a brand new session.
At SEJ, for instance, we now have the brief hyperlink “sejr.nl” area, which needs to be handled as the identical area – so we added it to our exclusion record.
Additionally, if in case you have subdomains and need to observe throughout subdomains utilizing the identical GA4 property, you must exclude your personal area from referrals to be able to maintain the identical session when customers navigate from one subdomain to your essential area.
7. Not Selecting The Proper Reporting Id
The next reporting identification choices can be found in GA4:
The excellent news is you’ll be able to change forwards and backwards between these choices anytime, and it’ll mirror in your customized exploration studies.
However I want to point out why it’s essential to decide on the precise choice in keeping with your enterprise case.
For those who don’t have login and person IDs in your web site, 99% of instances ought to go together with “device-based,” as a result of the opposite two choices might distort your conversions knowledge.
The reason being the person’s privateness. With Google alerts enabled, GA makes use of person IDs to trace customers throughout units, then matches them if they’re logged in to their Google service accounts on totally different units – and there’s a likelihood that person identification could also be uncovered.
In such instances, it hides person knowledge from the studies and fashions knowledge primarily based on person habits. Modeling of information can introduce some stage of inaccuracy as a result of it’s an estimation reasonably than an actual measurement.
With modeled and noticed choices, you’ll typically discover “Information thresholds are utilized” in your studies which have implications for knowledge accuracy.
You could attempt switching between these choices and see how your knowledge is altering.
For those who discover a major distinction within the variety of conversions between blended, noticed identities, and device-based, it might be preferable to make use of the latter choice.
Machine-based identification works equally to how Common Analytics monitoring works.
In conclusion, it’s essential to keep away from frequent configuration errors when establishing Google Analytics 4 to make sure correct and dependable knowledge assortment.
By understanding these potential pitfalls and taking the mandatory measures, you’ll be able to profit from GA4’s capabilities and derive significant insights to your web site or software.
Moreover, GA4 requires ongoing upkeep reasonably than a one-time setup.
Failing to repeatedly monitor and analyze your knowledge can result in missed alternatives and make it tough to determine and handle points on time.
Featured Picture: Solid Of Hundreds/Shutterstock