GA4 Server Side: 30% less purchase conversions

Hi!

  • We have set up Server GA4 tracking in parallel to the existing client side GA4 tracking
  • So there are 2x Google Tags running on web GTM
  • This is running since 3 weeks
  • usually after 2 weeks, our SST setups track more than the client setup

Results

Now if I look at the last full week, comparing SST GA4 vs Web GA4 property

events count

  • 24% less purchase for SST property
  • 70% less page_view for SST property
  • 30% less view_item for SST property

other

  • same amout of users
  • 60% less purchase value for SST property

Trigger and variables are the same for both, everything should be best practice.
Meta Conversions are fine
Google Ads Conversions are fine

I am now comparing Data Tag purchase and GA4 purchase with sGTM to google sheets, to see if there is a difference.

What to do now? Any Ideas?

Is there anything we should be looking for or test?
Is it because there are 2x Google Tags on client side, so maybe the second is not loading properly or too late?

I would appreciate your help a lot. :pray:

Similar Topic: GA4 server side only tracking 85% of purchases

Comparing GA4 performance in this way is bad practice and you can only by luck and very, very correct setup get normal results.
If you just add a new property and tags for it in parallel with the web - you will get bad results because GA4 is not designed to run multiple properties in parallel, especially if one of them is server-side and the other is client-side.

You may note that for using multiple GA4s in parallel there is a separate Rollup property functionality which is only available in the 360 version [GA4] About roll-up properties - Analytics Help

If you could just add a couple of extra google tags it would hardly be one of the 360’s features :slight_smile:

Also, here is an excellent article on how to try to correctly configure parallel operation of several GA4, you can estimate the number of nuances there: Build a GA4 Roll-up Property Using Server-side Google Tag Manager | by Markus Baersch | Medium

And in general there is no point in such a comparison. GA4 in server side format works as a proxy. Data is not changed in any way, nothing happens to it (unless you have specially configured it, which is usually very rarely done), requests are simply redirected further to GA.
So there is not much to compare here.

If you want a super safe transition - in GA4 client switch on JS managed cookies at the beginning, when after some time you are sure that everything is fine - switch to server managed cookies and activate the migration option. In this case everything will definitely go without problems if you don’t have problems with tag configuration.
If you have problems with unassigned/not set traffic sources when switching to server managed cookies, here is an article that describes the problems and solutions perfectly: Unassigned and not set traffic source in GA4 - Stape

1 Like

I removed the double-tracking and now it tracks about the same amount of conversions.
We deployed custom loader and I hope to see more conversions compared to before.
Will update!