A Casual Study of the Unsubscribe

A few months ago, I decided to change my personal policy on spam. For years, I just ignored spam, mostly because it was easier to delete messages than to follow the unsubscribe instructions on each of them.

Given the high profile hacks of the last 12-18 months and the nature of spam lists, i.e. lists of email addresses are sold without explicit consent in some cases, it’s difficult to know who has your email address, what they plan to do with it and how good their security is.

So, I started unsubscribing from spam as it arrived, and during this process, I’ve observed a wide range of options, which I’m ranking here for posterity. Unsubs are an odd use case because the motivation of the user is opposed to the motivation of the provider.

While legally, a bulk emailer must provide an unsubscribe function, it’s not in their best interests to do more than the absolute minimum. In fact, some of the implementations I’ve seen make it just confusing enough to persuade the user to stay, or at least, not leave before giving up more personal information.

Without further ado, here are the unsubscribe methods I’ve observed, from best to worst for discussion and humor purposes.


Without a doubt, the one-click, instant unsubscribe link is the best. I’ve found that some links are described as instant in the spam message, but are not in practice. This one functions as advertised, click it and you’re magically removed from the list. There’s usually a confirmation message and sometimes a timeframe listed for the unsub to take effect.

One-Click and Confirm

Sometimes “instant” means clicking the unsub link drops you on a confirmation page, i.e. the page wants to make sure you want to unsubscribe from the list and you haven’t hit your head or fat-finger-clicked the link by mistake. Because the benefits of subscription are so good, it’s likely that you made a mistake.

It’s one extra step, not a huge deal, but a little more annoying, especially if they go overboard on the benefits of their spam.

One-Click, Confirm and Tell Us Why

Quite a few unsubs have an optional step at the end of the process, like the break-up that won’t end.  I keep hoping to see the “It’s not me, it’s you” option, but alas, not yet. This one is obviously marketing, and it’s a little sneaky. With unsub function completed, there’s no reason to stay on the site, but this question, usually in form of radio buttons, takes just enough cognitive effort to make you question yourself.

I find myself plowing through this process, hoping to get it done as quickly as possible, and I know I answered that question a few times without even thinking.

Links That Make Me Work

Beyond confirming, some unsubscribes claim to be generic and require you to enter the email address you want to remove because there’s no way they could know what email address you mean, even though those links carry when look like unique tokens in their URLs. Don’t make me do work to unsubscribe, and definitely don’t pretend like you don’t know who I am.

Even worse is the lazy unsub link that actually doesn’t carry any tokens or identification. It’s just a generic unsubscribe page; I’ve found those usually also have the sneaky break-up question too.

Hall of Fail

Now, for the best of the worst.

I suppose the worst should be spam with no unsubscribe offered at all, which is technically illegal, but good luck enforcing that. These emails come from bots that scour the interwebs looking for email addresses, or so I assume, since they never seem to be solicited in any way.

Slightly better, but still highly annoying, is the reply with “Leave Out” or “Unsubscribe” in the subject. Or put differently, please confirm this is a real person and optionally provide more context about yourself in your signature. There’s never any confirmation that these work, and they feel a lot like bait.

I love the unsubscribes that tell you how long you should expect the “process” to take. It’s usually listed in days, sometimes with the hilarious caveat of “business days.” If only technology could automate the removal of a data point from a data store.

And finally, the worst and funniest unsubscribe I’ve seen so far: optionally send a letter. I kid you not. This spam message had a perfectly good unsubscribe link in the footer, and they conveniently offered a snail mail option, complete with a PO Box.

That made my day. Then, I felt a little sad that people might actually try to use that option.

Find the comments.




  1. I actually prefer “One Click and Confirm” to “One Click.” Unsubscribing from a list is a destructive action, and I prefer to have a consistent confirmation step for any destructive action – whether it’s the act of unsubscribing from a mailing list that I value, or the act of unsubscribing from a spammer’s list.

  2. @John: Sure, it’s supposed to be a destructive action, assuming they follow the letter of the law. I don’t really mind the confirmation step, but it’s not really accurate to call it one-click although some do 🙂

  3. I question whether bulk emailers and unsubscribers have interests that are that different. If the emailer makes it too hard to unsubscribe, the unsuccessful unsubscriber may just mark all future emails from that emailer as spam. I would expect email services to watch this. It would explain why some mail I want to receive ends up in the spam folder.

  4. http://www.ftc.gov/os/2008/05/R411008frn.pdf

    Section 7704(a)(5)prohibits the initiation of a commercial email message unless it contains three disclosures:
    (1)clear and conspicuous identification that the message is an advertisement or solicitation;
    (2) clear and conspicuous notice of the opportunity to decline to receive further commercial email messages from the sender; and
    (3) a valid physical postal address of the sender.

    I’ve long used the domain name where I sign up for things as part of my email address, including rude words when I think it is spam harvesting. I’ve been surprised at how little abuse there actually is, although recently I’ve been getting some fuckinlinkedin spam. Of course, I used to never reply to unsolicted links to unsubscribe, in the days when most of those were probably bait. I also use the trick of using html for the at sign to put a email on the web, though that failed for a while when google started helpfully pre-translating html to word and what-all.

    When I got a response to one complaint to linkedin, my email provider marked it as spam lol. It was a stupid response including how to surf safely, pretty much completely ignoring my complaint. Typical lame support.

    Some other fails:
    Lists that forget to blind copy recipients, and people try to unsubscribe with reply all.
    I saw one place do this with apparently a new list appended to the old one, causing a huge “I didn’t subscribe to this” “stop replying to me this is not my list!” … at least twice.

    Stupid lying republican and tea party lists – why would they even try? Then why would they even keep trying? All I can guess is they pay nefarious people by the email.
    Addresses I stopped using years ago (but still have forwarded) sometimes revivify with strange things, including obvious list merges that are complete nonsense.

    Different senders with the same subjects, how stupid is that? Be My Little Bunny, says the russian pediatrician who wants to meet me…

  5. Oh yes, spam on to usenet can make for some classic fail. Just a sample:


    And of course, the silliest ones put in a sig file that quotes a non-existent title in the US code saying that the spam is ok!

  6. @Brent: Hmm, if it’s impossible to unsub, then marking as spam is correct and technically illegal. I’ve only found a few that are impossible. I’d say those mailers don’t care either way.

    You might be giving spam filters too much credit. Marking something as spam shouldn’t be a terribly strong signal that applies to everyone. I’ve done that by accident, and personal definitions of spam vary wildly.

  7. @Joel: I like the use of profanity as a spam filter, kudos. Of course, one reason I mostly ignored spam for so many years is the old unsub bait. The law has cleaned that up quite a bit since the bad old days.

    I haven’t seen a cc-storm unsub for a while, but they are amusing. The anger that flows from some people is astonishing.

    Nice anti-spam examples, Crisco certification is great. I wonder if that’s fail or on purpose to avoid a lawsuit? Not that one would stick, just seems so obvious.

  8. @Jake, just because it’s technically possible doesn’t mean it’s worth the effort. The worst offenders are those that make you log in to unsubscribe. Seriously?!

    In my view, spam is unwanted email that you can’t easily get rid of.

  9. @Brent: Ah yes, I forgot to mention those; that feels like bait, very uncool.

    There’s a big difference between unsubscribing, which should remove your address, and marking as spam, which just hides the email from you and might result in some mitigation from your email provider or ISP.

    Not that unsubs are awesome, but I’d rather spend a little effort (in most cases) to do the work myself.

    Marking as spam relies on others do work for me.

  10. I don’t like the one-click option. Yes, for spam it’s ok, but for newsletters that I actually want not. Sometimes when seeing something nice I tend to forward the email to someone else. The one-click options in the footer gives the one I forwarded the message to the option to unsubscribe me.
    Although I forwarded the message, the receiver might think “what is this for spam” and click the unsubscribe button.
    Because of this, I like the “Links That Make Me Work” better. This person has to enter my email address to unsubscribe and in the above case, it will probably enter his own.

    Also, I am missing the “Log in to unsubscribe” one. If I find it spam, I probably dont’t know my password (anymore).

  11. @Anon: We have different workflows I guess. There are only two, maybe three, email newsletters that I read, and I’ve been ignoring spam so long that the one-click is nice for me.

    Forwarding a message like that sometimes trips a spam vector, and I don’t get much news by email, but I can see the problem for you.

  12. When iCloud was .mac and it gave me unlimited email alias – I used to use a different alias to subscribe to different ‘things’ and then when I got SPAM I could figure out who it was that sold/shared/had stolen my email address.

Leave a Reply

Your email address will not be published. Required fields are marked *