It would be really interesting and ultra-helpful to have A/B testing for metadata in terms of placement in YouTube SERPs for a particular search term. While CTR is an indirect measure of that (higher placement often equals higher CTR), but I think placement in search results could be more helpful.
My recommendation would be to switch the metadata at midnight and measure at some set time of the day (maybe the channel's high traffic hours just as a place far enough away from the midnight switch such that YouTube has had time to move/place it in the SERPs.)
My recommendation would be to switch the metadata at midnight and measure at some set time of the day (maybe the channel's high traffic hours just as a place far enough away from the midnight switch such that YouTube has had time to move/place it in the SERPs.)
