We have been getting the question “Why are my revenue numbers off?” quite a bit from clients in the last month, and we’re not surprised. Many brands delayed their GA4 migration until the last possible date, which was July 1st. GA4 is a big change, with entirely different definitions and attribution models. July was an eye opener.

There is a lot to unpack with GA4 and revenue reporting. Let’s start with the topline impact. Google support has shared directly with our team that a +/- 18% difference in reported revenue per channel is about what they expect, and we’re largely seeing that hold true across our clients. If you are seeing huge reported revenue swings which are not consistent across channels, you’re in good company.

There are three big causes for this which brands should be aware of: a) third party migration issues, b) changes in definitions, and c) new attribution modeling.

Third Party Migration Issues

GA4 conversions overall have had some bugs – even in some of the big, well-known platforms – and so your brand might be impacted by a third party integration that is not reporting numbers accurately at all. We’re not going to name names, but you should be speaking with any integration partners about the status of their GA4 conversion and the issues they’ve seen in reporting. This is a tough change so it’s not surprising, just something to keep an eye on and work together with your integration partners to figure out.

Changes in Definitions

GA4 is redefining just about everything – sessions, conversions, bounce rate, and more are all redefined and measured differently. We won’t list all the changes – Google support did a good job here – but brands should expect that they won’t be able to recreate all of their metrics for YOY comparison purposes. Brands should expect that revenue reporting will look different day to day and in different directions on each metric. Even changing the settings back to last click in GA 4 will not deliver the same results as before.

There is a lot of good reasoning behind the change, but it will make it tough to compare results for the next period. Our best advice is to do two things simultaneously: 1) embrace the new metrics for the value that they can and will bring to your program, and 2) recreate a few key metrics for the purpose of YOY reporting and making comparisons. Those metrics will still vary due to the change in definitions, so this isn’t a complete solution. You’ll abandon this second set in time, but you’re going to need them in the short term to help manage your business while you build history around the new metrics and definitions GA4 has implemented.

New Attribution Modeling

The biggest change for revenue reporting comes from the new attribution model. The default in Universal Analytics was last click attribution. Last click attribution is pretty easy to understand, even if we disagree with the picture it paints. It does exactly what it says.

The new default is the cross-channel data-driven model. That is a mouthful, but Google support does a reasonable job of explaining how it works here. The model is a machine learning model, which essentially means it optimizes itself for each brand, and it is completely unique to each brand and even to each conversion event. Because it can adapt to the brand and to the event, it allows for fractional attribution across the channels that GA4 can see.

This type of attribution across channels is great, except that it is a black box and brands cannot see how or why it is attributing clicks in any given instance. That makes it hard to apply those learnings to your future budget.

The Path Forward

There is no going back to Universal Analytics, so brands need to find a path forward. In the short term, understanding why reported revenue has changed so dramatically – and that it has done so for all brands at the point of transition to GA4 – will help smooth the path for productive reported revenue discussions. Recreating a few key metrics will help link to the past, but the real win will be in embracing the new metrics and aligning the organization around the new attribution model.