This message was deleted.

Hi Veronika 👋 The recommended practice here is to create a new experiment with the new variant. Adding a new variant to an experiment already in progress can lead to compromised result data.