Using SFBX backoffice, you can modify and update your notice settings. But you may want to compare the consent rate of differently defined notices to implement the notice generating the most consent-in.
Since the version 1.7.0 of the web CMP and the version 2.1.0 of the SDK Android and iOS, the SFBX backoffice allows you to lunch experiment campaigns in order to test and compare performance of different notice configurations (with or without banner, order of CTAs, user interface in one step or in different steps...).
You can create an experiment with 2 or more records running on the same time, on the same source. This various notices are called "Variations". Each variation will be exposed alternatively to users, according to your parameters.
An experiment can be implemented at anytime on a source.
While an experiment is running, you can consult the consent rate of each variation through the dashboard. But you can't modify the settings of the notice variations.
You can stop the experiment whenever you want. At this moment, you will have to choose one of the notice variations. The one you choose will be implemented definitively on the source, and your experiment stops. After that, you could modify the parameters of this notice if needed, or start an other experiment.