8

I'm using Exim as an MTA to send emails. Is it possible to get notified if an email goes into the recipient's spam folder?

psmears
  • 461
  • 3
  • 8
Abdessamad Elouarti
  • 93
  • 1
  • 7
  • 3
    Since the answers say no, probably your best bet is to measure how many people engage with your emails by tracking clicked links. In cases where there are significantly fewer clicks than usual, it may be because the mail ended up in the spam folder. – joeytwiddle May 24 '16 at 03:10
  • 14
    If you could do that, that would defeat the point of having a spam folder, because you'd just keep sending slightly different emails until one didn't go into the spam folder. – user253751 May 24 '16 at 05:41
  • Send a follow-up email and ask if the first one went to the spam folder :-) – MonkeyZeus May 24 '16 at 18:03
  • 4
    I'm voting to close this question as off-topic because it is about email in general, and not Unix/Linux specifically. It might be on-topic at [SU] (or possibly [SF]), but it probably won't get any answers that are different from the ones it has already gotten. – G-Man Says 'Reinstate Monica' May 24 '16 at 19:35

4 Answers4

14

No, you will get a "delivered" notice or an "error" notice. Once the mail is accepted by the remote end, you don't get to know where it goes after that. At least not on the MTA side of things.

One of the errors may be "rejected cause of spam" or "rejected because of SPF" or the like, but if your email is accepted, even to the spam folder, you will not get a notice. If the email is rejected by their server then your recipient will not get the email, even in their spam folder.

You may get an error of "Deferred" -- that may be because you are suspected of sending spam. This (the deferred status) will tell Exim to try again later. You may be able to get more information from that message. However, deferred is common, and normal, and not really an issue. Using it for spam warnings is very specific to the receiving end, and likely would not tell Exim do to anything but try again later.

Some services have "tricks" to see if a mail is marked as spam. Combinations of links, images, and maybe even javascript that can tell, in some cases, if your ending in the spam folder. But these don't work 100% of the time, and are more on the client (gmail, outlook, etc.) side then the MTA side.

Mark Stewart
  • 796
  • 1
  • 7
  • 20
coteyr
  • 4,280
  • 16
  • 24
  • 1
    Generally, SMTP is designed such that messages do not just vanish into oblivion (i.e. you're supposed to get an explicit refusal or bounce message). But SMTP is also designed under the entirely faulty assumption that spam does not exist, so this is hardly true in practice. Still, if you have DKIM and reverse DNS properly configured, and your IP address does not belong to a block known for spamming, it is unlikely that your email will disappear entirely (it may land in a spam folder, but it won't just get silently dropped). – Kevin May 24 '16 at 06:04
  • 3
    @Kevin unfortunately some mail hosts do silently bin email identified as spam — it's accepted at the SMTP stage but not delivered. – Stephen Kitt May 24 '16 at 07:08
10

No, what happens at the destination is visible only there (unless it bounces your email back to you).

Stephen Kitt
  • 411,918
  • 54
  • 1,065
  • 1,164
5

No, what happens after your email leaves your system cannot be traced unless you have access to the receiving computer, or if there is some spam filter program that gives feedback.

Such feedback programs would sent an email back asking for confirmation, so that automatic spam could be distinguished from real message sent by humans, but once bots were smart enough to answer this just generated extra mails and I haven't been asked for such a confirmation via email in over 15 years.

Bouncing emails is discouraged, if at all an email should not be accepted for delivery (if the address doesn't exist), but that that has nothing to do with knowing whether your message lands in the spam box. Spam should never be bounced as the sender in the mail header was unlikely to be the actual sender.

Anthon
  • 78,313
  • 42
  • 165
  • 222
  • if you're talking about challenge-response systems (aka user-level automated backscatter spam), I wrote procmail rules to deal with them which would **always** validate any challenge-response message sent to me. It's not my spam to deal with, it's theirs, and it was extremely offensive and objectionable for C-R users to expect ME to deal with THEIR spam - if they're going to offload their work on me, I'll just auto-accept it on their behalf. A valuable lesson on the risks of outsourcing. – cas May 24 '16 at 03:28
  • thnks for ur reply, I think u talk here about feedback loops, that helps to trigger recipient "complaints", so when a recipient mark the mail as spam, FBL allow senders to receive messages back from members who complained – Abdessamad Elouarti May 24 '16 at 08:05
  • @ABDESSAMADELOUARTI Yes you can see this as feedback loops. There are several problems with this, and generating an extra mail for each mail sent is one of them. Apart from that if A and B have both set up such a system with different mechanics, and I sent an email as A to B, then B might sent A a confirmation message, which A doesn't recognise as such, therefore auto-sending an email to B to confirm, which ...... ad infinitum – Anthon May 24 '16 at 09:50
3

There is no functionality that will tell you which folder your mail was delivered to. Even the tricks that tell you if the message was read, don't tell you where it was read from.

However, you can get a good idea from the major providers, gmail, yahoo, and microsoft, if you configure DMARC. This builds on your DKIM and SPF configuration, and can report back the number of messages received from your domain, as well as likely disposition. It will also report the IP addresses that are spoofing your domain(s).

BillThor
  • 8,887
  • 22
  • 27