Here is important news for marketers and digital product teams
It may come as a surprise to someone, but Google Analytics 4 will be removing first-click, linear, time-decay and position-based attribution models from its attribution reports.
Starting in May, newly created Google Analytics 4 properties will no longer be able to select these models. In September, Google will sunset these models for all other GA4 properties. Cross-channel data-driven attribution, cross-channel last click, and Ads’ preferred last click will continue to be available.
Data-driven attribution in Google Analytics 4 is Google’s preferred option, and Google decided that’s now your best option to evaluate marketing channels. Data-driven attribution in Google Analytics allows Google to measure its own Google Ads performance. It seems that this is not a very independent attribution that is affiliated with Google.
Of course, you can calculate and use your own attribution based on raw data and use a ruled-based or machine-learning approach. So it will be another reason to collect and use raw Google Analytics 4 events in BigQuery.
Which attribution model is the best?
When it comes to comparing attribution models in Google Analytics, the question is often, which attribution model is the best. I would say that this is not a very correct question.
The truth is that no one model is absolutely better than the others, as each model illustrates one aspect of the customer journey in reality. In order to compare multiple models, digital professionals need to use multiple models because they see the goal.
How to learn more about attribution
We have already started talking about attribution models here
Coming soon, we’ll be sharing the following topics on how to use an independent attribution approach for any marketing channel instead of Google-affiliated data-driven attribution in Google Analytics 4.
- Single-touch rule-based attribution models
- Multi-touch rule-based attribution models
- Post-view (view-through) vs Post-click Attribution
- Machine Learning Attribution approach
- Shapley Value machine learning attribution
- Markov Chain machine learning attribution
- BigQuery ML Attribution with prediction modelling