This is likely due to the change in Google Analytics’ reporting system. They moved from “Universal Analytics” (UA) to “Google Analytics 4” (GA4) which included a significant change in the way they count things like pageviews. It’s not exactly that the numbers we had before were inaccurate, nor that the new numbers are inaccurate, it’s that they’re being counted in a different way. Here are 2 major differences:

1. UA tracked each device as a user; GA4 tries to track the same person across multiple devices (e.g. phone, tablet, laptop, desktop computer)
2. UA would count the same person visiting on different days as different users; GA4 will track the same person visiting multiple times as a single user in a given report. So if you ran a report for 2 days, the same person visiting twice would count as 2 users in UA, but only 1 user in GA4. Now expand this out to an annual report, and you can imagine how the numbers can become much lower if you have a fair number of repeat visitors.
3. GA4 may be doing a better job of honoring browsers’ Do Not Track requests, plus over time more people may be using extensions that block 3rd-party cookies and traffic, like Google Analytics.

To demonstrate how point #2 above can lower our numbers, here are some tables comparing longer-term reporting periods with sums of shorter-term reporting periods:

Monthly vs. Annual

Comparison:

• Total users in single report from Jul 1, 2023 to Jun 30, 2024: 24,548
• Difference between reporting methods: 1,190
• % more users when broken down by month: 5%

Daily vs. Monthly

Comparison:

• Total users in single report from Oct 1, 2023 to Oct 31, 2023: 2,579
• Difference between reporting methods: 580
• % more users in October 2023 when broken down by day: 22.5%

For anyone who’s interested, here’s a bit more information about the differences between UA and GA4:

• [GA4] Differences in user counts
• GA4 Users vs UA Users
• The trouble with Users in GA4

I hope that helps explain some of the difference. I am surprised the number dropped in half – that is a pretty substantial change. If you’d like, I can spend some time delving into that in more detail to see if there’s something going on, but my hunch is that it’s mostly attributable to the change in the way Google Analytics records and reports its data. I’m not an expert at Google Analytics; if you know anyone who’s good at analyzing Google Analytics reports, we could invite them to take a look and see what they notice.

</Michael>

Michael McKelvey

Office for Mathematics, Science, & Technology Education