Note important prerequisite:
If you're going to use this article, it is because you have discussed implementing a custom AB test solution with your Customer Success manager at DataMilk. If you have not discussed a custom AB test solution with the DataMilk team, this article does not apply to you and it will not yield expected results.
Step 1 - Create custom DataMilk Report
Log in to GA (client should have sent you the access invite link)
Go to Customization menu on left panel
→ click Custom Reports
→ click + New Custom Report
You’ll land on this screen:
In the General Information section, replace the report ‘Title’ to DataMilk Report
→ Click the blue +add metric button in the Metric Groups sub-section
You’ll be given a dropdown menu with a search bar - type-search for the following metrics and click them once to add one by one:
- ‘Ecommerce Conversion Rate’
- ‘Sessions’
→ Click the green +add dimension button in the Dimension Drilldowns sub-section
You’ll be given a dropdown menu with a search bar - type-search for the following metric and click it once to add:
- ‘Event Label’
- 'Event Category'
- 'Event Action'
→ Click the gray +add filter button in the Filters section below
You’ll add 4 filters. Note that when you type to search for a term, a dropdown search list should populate for you to select.
Add filter 1:
→ Click +add filter button
→ type/select ‘Event Category’
→ leave first button as default 'Include'
→ leave next button as default 'Exact'
→ type/select 'DataMilk'
Add filter 2:
→ Click +add filter button
→ type/select ‘Event Action’
→ leave first button as default 'Include'
→ leave next button as default 'Exact'
→ type/select 'A/B'
Add filter 3:
→ Click +add filter button
→ type/select ‘Event Label’
→ leave first button as default 'Exclude'
→ leave next button as default 'Exact'
→ type/select 'optimized'
Add filter 4:
→ Click +add filter button
→ type/select ‘Event Label’
→ leave first button as default 'Exclude'
→ leave next button as default 'Exact'
→ type/select 'original'
The filters list should look like once the above steps are complete:
Your final DataMilk custom report should look like this from the 'Edit' screen:
→ Click the Save button at the bottom
Step 2 - Interpret the Results
Open your custom DataMilk Report in GA
Your report should look something like this if Step 1's instructions above were completed correctly:
- Adjust the date range from the menu at the top right of the screen accordingly.
- Calculating Uplift: Simply calculate the percentage increase (or decrease) from 'original' to 'optimized'. In this example, we calculate a +12.74% conversion rate uplift.
The Conversion rate and Sessions count you see in GA will be similar and not exact to what is you see reported in your DataMilk dashboard. Please review this article to learn why: What to expect if I compare GA to DataMilk.