Ken Magill posted today on "Why Fully Confirmed Opt-in Sucks." It's definitely worth reading, and I hear where Ken's coming from.
To "lose a subscriber" through their failure to confirm, that can really hurt when a list is pretty small. I should know -- I do know this myself -- because I managed the email list for my friend's wonderful jazz club in St. Paul, Minnesota, from late 1998 through mid-2006. (That would be the Artists' Quarter, by the way, and you should definitely go there next time you're up in the Twin Cities. Tell Kenny and Davis that Al sent you.)
For the AQ email list, I used COI from the start. It wasn't necessarily a political statement. It was born of using the tools I had handy. I had previously written a confirmed opt-in list management tool myself, so that's what I used.
After about five years in, I think we had about five thousand confirmed subscribers. It grew pretty slowly but regularly throughout that time. A local jazz club's marketing efforts aren't necessarily like big business. Even so, eventually we hit a critical mass where the email messages became more important than monthly postal mailings of the club's calendar. The emails clearly drew patrons into the club and back to the club. The way we knew that is because whenever I would accidentally typo a cover charge in an email message, it was usually me publishing it as lower than the actual cover price, and a group of folks would always complain upon reaching the club's front door only to find that the door man wanted more for entry than it said in the email message. (Side note: Typos -- I never realized what a quick and easy measure of success they can be.)
Yes, occasionally somebody couldn't figure out to click a link. That's something you'd see less of today, I suspect, but back then it was no more than a handful of people per year who would reply to the confirm email or otherwise contact me, complaining that they weren't receiving the emails or didn't know what to do.
Somewhere, I've got the final signup logs, showing all the people who didn't confirm. Sadly, I couldn't find the data to share it with you (and it would be pretty old by now, wouldn't it?). But, my recollection is that what unconfirmed addresses were submitted were mostly garbage, primarily typos. And the occasional fraudulent subscription attempt. Being a loud and well known anti-spammer meant that people would occasionally try to indirectly cause problems for me by causing problems for my friends -- that's where that kind of thing came into play.
Anyway. Long story short, COI made sense in my specific use case, and wasn't problematic.
You've got random marketers here and there yelling about how you should never use COI (or even respect engagement), and you've got Ken's well considered article on his specific use case, so I thought I would share mine as well.
In closing, let me tell you what I told Ken over in the comments on his article: There's a certain class or certain type of marketer basically running around flailing their arms and yelling about how best practices are stupid, advice suggesting caution is stupid, we see other people do it wrong, why can't we do the same? Well, my response to that is this: Not every poor mailer gets blacklisted -- just like not all poorly built buildings fall down. That doesn't mean the caution recommended and expertise provided by your local neighborhood deliverability expert was wrong. It means you got lucky. Not everybody gets lucky every time, and some of us think it might be wise to build in some safeguards, since some of us think they're better than luck in the long run. Back to the building analogy -- I can't stop you from building it faster and higher, unsafely, but you also can't act surprised if and when it comes crashing down. Because I've seen it happen a couple times a month for the past ten years. Some months more than that, some months fewer, but it all averages out.