So, say we had one button to opt-in to messages sent out in two very specific situations. If each one of those three points got a 50/50 response, already 87.5% percent of people are at least somewhat dissatisfied with the implementation. It’s hard to call that anything but a failure.

Confronted with math like that, the natural instinct is to shut up and keep your head down.

(Keep in mind that most of the specific situations that would poll really well are individualized ones that we already notify about: your balance is low or zero, your domain is about to expire, we detected somebody trying to hack your WordPress blog, etc. Generic broadcast messages are much more balanced, as evidenced by the person who says they’re on freebsd6; quite a lot of people don’t care what we do as long as it don’t break their stuff.)

However, I also think the three comments requesting similar-but-not-identical things illustrate the fundamental problem with broadcast emails. Each of these three messages want to receive email updates, but each wants email updates about a different subset of things. Two say they think it should be opt-in. One sounds like it should be opt-out. No one solution meets the criteria of all three people, and that’s just three people.

The other problem is that every email sent that a person doesn’t care about reduces the amount of attention they’ll pay to future emails. We purposefully don’t email people very much so that when we do, they know it’s important.

So this is one of those rough areas where there’s nothing we can do that will satisfy everyone, and I’m sorry for that. So in the future we’ll do something basic and middle-of-the-road, like offer a more general option to receive blog posts by email. (Actually, probably notification of the posts.) But I don’t have any timeline for when that might be implemented. 🙂

As outlined in the post, we are sending email messages upon completion of the migration, since our system is able to analyze the site during the migration and try to spot trouble. Thus that’s the point where our system can generate messages for people that contain specific information relevant to them. What we are finding with that approach is that the number of sites for which this migration requires any action at all isn’t just less than 1%, it’s running about 1 in 1000. And in some of those cases, the action needed was due to a (now fixed) glitch on our part with the migration. So we do feel the decision not to send out a mass email is the correct one, as in almost all cases it would be repeating information that people already know about something they don’t need to do anything about.

In any case, I understand your point of view and I’m sorry we aren’t able to accommodate everyone’s individual contact preferences during unique situations like this. In the future, we would like to offer a more general option to receive blog posts by email, but I don’t have any timeline for when that might be implemented.