|
Hi all, There's an interesting thread on beta@ where Mark has been discussing a mechanism whereby certain email providers will report (Feed Back) to a list service such as Groups.io when a user marks a list's message as Spam. Apparently the email provider expects their user to be automatically unsubscribed from the list as a result. /g/beta/message/10753Does anyone here who uses one of the cited email services (AOL, Yahoo, and MSN) know about this mechanism? And in particular can you cite documentation about it in the service's help pages, user forums, or elsewhere? I have a feeling that this topic is likely to become something we should cover in the Advice to Group Members section of GMF's wiki. Mark has turned on the Feature for AOL addresses, says he will likely do so for the others too. So I suspect there will be ousted (and outraged) members banging on our groups' doors to find out what happened. I'd like to have some information here which fellow mods can point to that will help answer their members' questions and maybe even provide useful service-specific guidance on how to avoid being ousted again. Shal
|
Shal, This is exactly what I'm afraid of. I've already been advising members (in strong terms) that they should not use an AOL email address for the group, or there are no guarantees about them getting all messages. The new setup could provide a way to emphasize that, but it needs to be done in such a way that Groups.io is not blamed. I want to be preemptive about this before the ousted and outraged members come banging at our doors. Would it be legitimate to tell members right now that "if AOL marks a message from Groups.io as spam, you may be automatically unsubscribed from the group"? That puts the blame squarely on AOL's shoulders, where it belongs. I assume it's correct to say that AOL is responsible for the marking of Groups.io messages as spam. But is it? J
|
I found this in AOL's help pages. ?
How do I unsubscribe from email newsletters that I receive?
Learn how to unsubscribe from newsletters and daily deals email.
Is your inbox getting cluttered with newsletters and daily deals? If you want to stop receiving these newsletters, we can send an unsubscribe request to the sender on your behalf.
Here's how:
1. Find the newsletter or daily deal that you'd like to unsubscribe to and click the?Spambutton at the top of your Inbox.
2. If it's a newsletter that we can help you unsubscribe to, you'll see a box with anUnsubscribe?button. If you're sure you want to stop getting these newsletters, click theUnsubscribe?button.
?
That's it! Your unsubscribe request has been sent. If you want to re-subscribe to this newsletter you'll need to contact the sender and sign up for their newsletters again.
Note:?Just because we sent the unsubscribe request, the sender may still ignore it and continue to send you unwanted mail. Don't worry, if they continue to send you their newsletters, we'll send it direct to your Spam folder where it will be out of sight (and soon deleted!).
toggle quoted message
Show quoted text
-----Original Message-----
From: Shal Farley <shals2nd@...>
To: GroupManagersForum < [email protected]>
Sent: Fri, Aug 19, 2016 9:09 pm
Subject: [GMF] Information about spam FeedBack Loop (FBL) mechanisms
Hi all,
There's an interesting thread on beta@ where Mark has been discussing a mechanism whereby certain email providers will report (Feed Back) to a list service such as Groups.io when a user marks a list's message as Spam. Apparently the email provider expects their user to be automatically unsubscribed from the list as a result.
/g/beta/message/10753
Does anyone here who uses one of the cited email services (AOL, Yahoo, and MSN) know about this mechanism? And in particular can you cite documentation about it in the service's help pages, user forums, or elsewhere?
I have a feeling that this topic is likely to become something we should cover in the Advice to Group Members section of GMF's wiki. Mark has turned on the Feature for AOL addresses, says he will likely do so for the others too.
So I suspect there will be ousted (and outraged) members banging on our groups' doors to find out what happened. I'd like to have some information here which fellow mods can point to that will help answer their members' questions and maybe even provide useful service-specific guidance on how to avoid being ousted again.
Shal
|
J, This is exactly what I'm afraid of. I've already been advising members
(in strong terms) that they should not use an AOL email address for the
group, or there are no guarantees about them getting all messages. If Mark is right about AOL's expectations, this mechanism should improve delivery for most members. The theory is that AOL has been blocking some Groups.io messages at least in part because Groups.io hasn't been respecting the feedback about members who mark messages as spam. The new setup could provide a way to emphasize that, but it needs to be
done in such a way that Groups.io is not blamed. Mark hasn't indicated how many AOL users this might affect. But possibly it is a small percentage of any group's AOL-using members. The rest, hopefully the great majority, might see things getting better. At any rate the unaffected members wouldn't have anything new to blame on Groups.io, only the ousted members would. I want to be preemptive about this before the ousted and outraged
members come banging at our doors. Would it be legitimate to tell
members right now that "if AOL marks a message from Groups.io as spam,
you may be automatically unsubscribed from the group"? No, I don't think so. My understanding is that the volition here is on the part of the user, not AOL. That is, the feedback happens when the user marks a group's message as spam, not merely because a message was automatically directed to the user's spam folder. My only concern about this understanding is Mark's report that some people who were unsubscribed in the past claimed that they had not marked any messages as spam. One possibility is that those people failed to remember doing so, another possibility is that there's a more implicit way one gets reported. One question I'd have is what happens if a message is delivered to the spam folder, and the user deletes it from there rather than first designating it as not-spam. Does that constitute an implicit confirmation (marking) of the message as being spam? If so that could account for the users' belief that they had not done such marking. I assume it's correct to say that AOL is responsible for the marking of
Groups.io messages as spam. But is it? AOL is responsible for delivering messages that it deems to be spam to the user's spam folder. But what Mark said is "when their members mark emails as spam" -- that would leave out the automatic discrimination. That terminology is consistent with the usual behavior of content-based spam filters: usually the filter is only adjusted when the (human) user marks a message as "spam" or "not-spam" in contradiction to the filter result. If the filter adjusted itself based on its own results it would become "over trained", which usually makes such a filter worthless at predicting its user's preference. Shal
|
On Fri, Aug 19, 2016 at 08:04 pm, Shal Farley wrote: My only concern about this understanding is Mark's report that some people who were unsubscribed in the past claimed that they had not marked any messages as spam. Shal, That is my MAIN concern. Nobody who's subscribed to my group marks their emails from here as spam. What typically happens is that someone claims not to have been getting the emails, and then they look in their spam folder, and lo and behold, there they are. This to me says that AOL is the guilty party in mistakenly marking messages as spam.? Would it be correct to tell members that if they OR their email client marks a message from Groups.io as spam, they will be automatically unsubscribed? J
|
Kate, I found this in AOL's help pages.
How do I unsubscribe from email newsletters that I receive?
... Excellent, Thanks! For the record the URL appears to be: In general, and for this project, I'd prefer to have just a relevant snippet of the article together with a link to it. That keeps us clear of any copyright claims the author might have, and it gives us a way to cite the material going forward. In this case though I doubt AOL will have any concerns about a quote of their publicly available help text. And wow. I'm a bit taken aback that their mechanism is to have the member explicitly use the Spam button first rather than supply a more appropriate "Unsubscribe" button upfront. In my mind the two should never be conflated like that. It makes me want to shout in AOL's face: a list you no longer wish to hear from IS NOT SPAM. That's so wrong. At least the confirmation pop-up says Unsubscribe, but by then it is in some sense too late: the user has already been influenced to think of list messages as "spam" simply because they no longer wish to subscribe. Shal
|
J, That is my MAIN concern. Nobody who's subscribed to my group marks their
emails from here as spam. What typically happens is that someone claims
not to have been getting the emails, and then they look in their spam
folder, and lo and behold, there they are. So far no harm, no foul - at least as far as the feedback mechanism is concerned. The member may have been disappointed to find the messages in their spam folder rather than their inbox but presumably there's a means to mark the messages as "not spam" (which may automatically move them to the members inbox). This to me says that AOL is the guilty party in mistakenly marking
messages as spam. AOL is guilty of delivering the messages to the spam folder, but I'm fairly convinced (especially now seeing Kate's find) that delivery there would not of itself cause a feedback report to be sent to Groups.io. AOL's help page strongly implies that the user's action, clicking Spam and then Unsubscribe, is what causes the unsubscribe request (feedback report) to be sent. Would it be correct to tell members that if they OR their email client
marks a message from Groups.io as spam, they will be automatically
unsubscribed? I still do not think so, as described above. It really would not make sense for AOL to send a feedback report to Groups.io simply because the filter placed a group message into the user's spam folder. Shal
|
Ok, thanks, Shal. I think I'm starting to get a handle on this. J
|
On Fri, Aug 19, 2016 at 08:04 pm, Shal Farley wrote: One question I'd have is what happens if a message is delivered to the spam
folder, and the user deletes it from there rather than first designating it as
not-spam. Does that constitute an implicit confirmation (marking) of the
message as being spam? If so that could account for the users' belief that
they had not done such marking. I caught yahooMail doing exactly that. Users typically don't care to use the Not Spam button. I can believe that some email providers do that even if the message in Spam folder was not deleted by the user, but expired. :( On Fri, Aug 19, 2016 at 08:45 pm, Shal Farley wrote: AOL's help page strongly implies that the user's action, clicking Spam and
then Unsubscribe, is what causes the unsubscribe request (feedback
report) to be sent. Unsubscribe request is not the same as feedback report. An IP-address owner can subscribe to feedback reports. The how-to is in Help pages for postmasters, not for users. Unsubscribe button appears if the email provider found a standard line in the message header (List-Unsubscribe). BTW, groups.io digests haven't List-Unsubscribe line in the header.
|
So, the bottom line is that AOL makes it a point to teach their users to be
even bigger idiots than they were when they signed up for AOL
Sent: Friday, August 19, 2016 22:35 PM
Subject: Re: [GMF] Information about spam FeedBack Loop (FBL)
mechanisms
?
I found this in AOL's help pages.??
?
How
do I unsubscribe from email newsletters that I receive?[excess quote and long quote trimmed by moderator]
|
Lena, I caught yahooMail doing exactly that.
Users typically don't care to use the Not Spam button. This would seem to be a problem for any list service, with users of such email providers. If the list service responds to the spam feedback by unsubscribing the member, but the feedback was generated by a seemingly harmless action (deleting a message from the junk folder rather than not-spam'ing it first), that will be a bad user experience. It's hard to say who the user will blame for it, which is why I suggested that Mark make the message to the member very explicit that the mail provider initiated the action. I can believe that some email providers do that even if the message in
Spam folder was not deleted by the user, but expired. :( Alas, I too can believe that a service would mindlessly treat both kinds of deletion the same. But yikes, that's a much worse user experience: they get unsubscribed "out of the blue" when a message they never saw ages out of the spam folder. Unsubscribe request is not the same as feedback report. An IP-address
owner can subscribe to feedback reports. The how-to is in Help pages for
postmasters, not for users. Unsubscribe button appears if the email
provider found a standard line in the message header
(List-Unsubscribe). Ok, so if the header field is there, and the unsubscribe button clicked, then the list manager gets back the mailto: they put in the header field as distinct from a feedback report. But whether the list gets an unsubscribe message or a feedback report isn't really the issue here. The concern that caused Mark to take down the FBL mechanism a while back was that many of the affected users wrote to him claiming not to have unsubscribed. I was wondering if those users simply forgot or misunderstood their action that caused the unsubscription. Unfortunately you've confirmed, at least in the case of Yahoo Mail, that the member's action could have been a simple deletion, which they had no reason to believe would result in unsubscription. Or worse, we both suspect that it could happen without any action on the member's part. And it is that latter case, where simply having a message delivered to the spam folder and later expire might cause a feedback report, forces me to reverse my answer to J - it just might be possible that merely having the provider determine that a message is spam could cause a feedback report and result in unsubscription. So the advice against using such email providers isn't wrong. It's is either that or be vigilant about checking one's spam folder and marking group messages as not-spam. Of course, that's a good idea with any email provider - if for no other reason than that you might have missed those messages and marking them not-spam may eventually help reduce the chances that future group messages will end up in the spam folder. BTW, groups.io digests haven't List-Unsubscribe line in the header. Oh. Have you reported this to Mark? That might be a contributing factor also; I don't have any way to know how many of the people who complained might have been on Digest but he might. Shal
|
On Sun, Aug 21, 2016 at 05:06 pm, Shal Farley wrote:
groups.io digests haven't List-Unsubscribe line in the header. Oh. Have you reported this to Mark?
I already bothered him too many times, I hesitate. :)
|
I would think that having a "List-Unsubscribe line in the header" would be an important item for services like YahooMail or AOL. It should also help Mark prevent problems with other services that may implement this tactic later. I'd "bother" him about it. ;>)
Duane
|
Shal Farley
Lena
Duane