Widget a/b testing control group and view-based attribution

Several exciting changes were released for widget analytics.

Control group in widget experiments

You can now toggle on a control group for your experiments. This allows testing the performance in terms of goal hits and revenue for both visitors who see a widget and those who don't.

View-based widget attribution

You can now configure the widget to attribute goal hits (or purchases) to widgets based on a view, instead of needing the user to perform a conversion event on it.
This will help you analyze the impact your uninteractive widgets, like promo bars, have on your revenue and will also give more flexibility when running experiments.