Customize attribution models in GA

1. Determine the strategic uses for which you’d like to use attribution modeling, such as making channel marketing decisions or understanding the buyer’s journey.

Navigate to Conversions > Multi-Channel Funnels > Assisted Conversions to view the Assisted Conversions report, one of the most useful reports to start looking at before you set up custom attribution models in Google Analytics. It shows how different channels work together to build conversions and can give you insight into what a custom setup might look like.

2. In Google Analytics, navigate to Conversions > Multi-Channel Funnels > Model Comparison Tool to view your existing attribution model comparisons.

Make sure you actually need a custom model in the first place as often the built-in ones can answer your business questions. Click on the drop-down and go through the existing attribution models: Last Interaction: gives credit to the channel that produced the final visit before the conversion. Last Non-Direct Click: gives credit to the channel that produced the final visit other than a direct tying in of your website name. Last Google Ads Click: gives credit to the last visit from a Google Ad. First Interaction: gives credit to the channel that produced the very first visit to your website. Linear: gives equal credit to channels across all visits before a conversion. Time Decay: gives increasing credit to channels that are closer to the visit that produces the conversion. Position Based: gives more credit to channel producing the very first visit, and the visit that produced the conversion.

3. Select Create custom attribution model from the drop-down list and select your model parameters.

Give your new model a name. A common naming convention is: {{Person who set this up}} - {{Baseline Model Type}} - {{Lookback window Credit}} - {{date}} Choose a baseline model. Select on for the lookback window and set your desired amount of days. The maximum is 90 days. Consider enabling Adjust credit based on user engagement and choose time on site or page depth. These don’t always translate to success, so look carefully at your engagement metrics before adding credit to these. For example, time on site could mean users aren’t finding your content. Add Custom Credit Rules to access many of GA’s dimensions, like Default Channel Grouping, Source, Source/Medium, and Targeting Type. Set up multiple custom credit options, then choose the level of credit you want to give each one. The default level is 1 and generally works well for most models. If everything gets a point, you don’t have to remember what you gave what and this becomes an attribution standard.  Save your custom attribution model.

4. Compare your new attribution model against the pre-built models available in Google Analytics to analyze its differences on the Model Comparison Tool page.

Select different goals to understand how the attribution affects the various conversions. If you are running Google Ads, Google Analytics makes it easy to look at all your campaigns right inside the model comparison tool.

5. If you’re using 3rd party tools such as Supermetrics and Google Data Studio, note that the Multi-Channel Funnels report is specific to Google Analytics and will not pass data via the typical GA API.

Data in the Multi-Channel Funnels report uses a different API to everything else in Google Analytics. If you are using something like Supermetrics to bring in the GA data into Google Data Studio, you will need to look for the right connection or API to use, the built-in GA connector to GDS does not bring in any of this data.