We don't know an actual conversion rate because the email blast was sent to our entire user base, and many will not be Chrome users, etc. I suspect, though, that we'd have to have many times the user base we have now to do a true cohort test of it. Even then, something like this is very qualitative, right? What if more people wrote love letters or reviews from one group, but the other group spent significantly more time writing much more enthusiastic material?
