Which evidence do you retain after a controlled test?

For this type of project I would keep the tier boundary explicit, use one contextual target link, and compare verified-domain coverage after the normal verification window.

For more info about campaign quality control have a look at the internet site.