- Digital Analytics Edge
- Posts
- Trust Issues with GA4: Data is neither Accurate nor Precise.
Trust Issues with GA4: Data is neither Accurate nor Precise.
Digital Analytics Insights, Jobs, Candidates and more...
Anil Batra
December 13, 2023
Table of Contents
One Insight You Can Apply Right Away
In this section, I share my thoughts based on my conversations with experts, clients, and students. These are not tips and tricks related to tools but are insights that will help you in your career.
GA4 is ruining Digital Analytics. Data is neither Accurate nor Precise.
GA4 keeps making changes and everybody tries to interpret what those changes mean.
Whenever a change is made, some people think it is a good thing, and others think it is a bad thing.
While people debate the merits of the change, GA4 makes another change, and the debate cycle continues.
This has been happening since Google Analytics announced GA4.
I was one of the early adopters of GA4 and liked the data model as it provided a lot of flexibility. However, the GA4 team mishandled the whole process, which led to a lot of disappointment. But there is only so much you can complain about when the product is free.
Free only works if it is acceptable and you can trust the product. This is where Google Analytics failed.
GA4 is ruining Digital Analytics.
In my Digital Analytics program, I teach that Digital Analytics data is never accurate but it is precise.
When something is accurate then you can be sure that you have got the data that is very close to the exact numbers. However, with any tracking tool, you can never get the actual numbers of users, visits, page views, etc. on your site/app. The reasons for that are cookie blocking, cookie deletions, ad blockers, firewalls, etc. So Web Analytics tools like Google Analytics were never accurate and never will be. Which is fine, we accepted it and worked with percentages and rations instead of actual numbers.
When something is precise then it will always produce a similar result all the time. even though it might not be accurate you can be sure that it is going to produce similar results.
That’s what Digital Analytics tools provide us, precise results. So instead of talking about actual numbers we use directional improvements to determine the performance of our site, app, and marketing.
However, with all the changes in GA4, I can tell you that it is not precise either.
First, we moved from Univeral Analytics to GA4, there was a huge difference in numbers between UA and GA4. Even directional changes did not make sense but we carried on because it is a free tool. Many “experts” came to their conclusions without knowing the inner workings of GA4.
Then GA4 applied thresholding, thus eroding the trust in the numbers presented in GA4. Unlike sampling, which takes a sample of the data, thresholding just withheld data. That means you only get a partial view of your data. The thresholding is mainly due to “Google Signals” enablement but there are other factors too. Most recently, Google Analytics removed thresholding as part of report identity and provided a very vague statement about its impact on the reports. (If you are not familiar with thresholding then read, https://www.optizent.com/blog/ga4-thresholding-applied-what-is-ga4-data-threshold-and-how-to-resolve-it/)
In a nutshell, you can’t trust the numbers in the GA4 interface because GA4 presents different numbers based on the options you choose and other "behind-the-scene” changes the Google product team makes.
I think things are going to get worse with GA4, as I don’t think the GA4 product team understands digital analytics. They make random decisions without talking to customers and people who are dealing with various client scenarios.
So what should you do?
If you are using GA4 and don’t see any thresholding applied then continue using it as is but keep an eye on the changes GA4 is making as they will impact your numbers.
If you are using GA4 and see thresholding in your reports then identify the cause and try to remove thresholding. If you need help then my team at Optizent.com can help you identify and fix those issues.
I also recommend using GA4 as a data collection tool and using BigQuery for any reporting and analysis, as BigQuery will not have the same thresholding issues as the GA4 reporting interface.
Finally, if you can’t keep up with GA4 then take a look at other tools. There are several tools in the market by my personal favorite is Piwik PRO. We do provide training and consulting services to help you move to Piwik PRO.
Two Free Educational Links For You
People Available For New Opportunities on AnalyticsHire.com
A lot of people around the world are looking for new jobs. Many have been laid off due to a bad economy. In this section, I want to showcase those who are looking for a job. If you have a job opening or can help with introductions then feel free to contact them. Click on the name to see their details and contact information. Also, if you are one of those who is looking for a new job then add your listing for free on AnalyticsHire.com and I will include your profile in the next newsletter. I am also happy to make personal introductions to the hiring managers and recruiters.
15+Job Listings
Here are 15+ Digital Analytics job listings that our team has hand-picked.