6 ms·

Do-not-reply

43 points by numlocked 17y ago

RyanMcGreal 17y ago
QFT: Using a do-not-reply address to communicate ... sends a message that the organization's time is more valuable than yours. Though there are legitimate exceptions - I'm thinking, for example, of an opt-in automated status update that has the sole purpose of sending you a piece of information to which you would have no particular need to reply.
- Torn 17y ago
  Exactly this. There's plenty of legitimate one-way uses of email communication. When I'm getting confirmation emails from Amazon after having bought something, for example, I do not expect to be able to reply to the machine sending them out -- it doesn't make much sense. I'll go further with my argument: the ability to reply via email doesn't guarantee my communication will be dealt with in an appropriate or timely manner. Far better, I think, to go to the support section of a site and use a guided process to get my query to either answer the question myself or to have it directed down the appropriate channel.
  - prodigal_erik 17y ago
    > the ability to reply via email doesn't guarantee my communication will be dealt with in an appropriate or timely manner. If they're too incompetent to handle email, I won't expect timely or appropriate action from them no matter what I do, and I'll regard doing business with them accordingly.
    - Torn 17y ago
      Having a single email address at the top of an email, necessitating parsing or eyeballing the content to understand the intentions and then forwarding it on to the appropriate people is surely more work, and less efficient, than the alternatives. It's conceptually easier to streamline support via a guided on-line process or knowledgebase type repository. Users can often answer their own questions, or if not they can at least be pointed at the appropriate department. I'd go as far as to say a large retailer would appear more incompetent if my only access to support was via a single email address.
      - numlocked 17y ago
        I completely agree if the email was the only point of access, but forcing customers onto a site is a classic case of forcing customers into a work flow that makes sense for the organization instead of providing tools to work with the customer's natural work flow.
dpcan 17y ago
False. 1) Larger companies have servers dedicated to sending out mass emails and the email address has to resolve back to the sending server properly or the spam filters kill it. 2) They need you to click the links in the email for analytics. Learning what offers get the most traction HELPS them AND YOU so they deliver relevant content in the future. 3) If you use a "support" or "sales" email as the sender and then send the mass email to 100,000 people, you are going to get 20,000 bounce-backs, littering your support email channel and support software with garbage. 4) You can't have people replying with "don't email me again" because those will get missed. Again, you can't have a large list and realistically go through all the bounce-backs looking for legitimate unsubscribe messages - the links IN the email need to be used for this. It's not fair to compare email campaigns to calling a company on the phone. Apples and Oranges.
- lsb 17y ago
  But these bouncebacks are predictable, no? There are several key words in them that identify them as bouncebacks, no? So make a spam filter for them, and it'll work great, especially because the mail programs aren't trying to evade spam filters with their messages.
  - dpcan 17y ago
    Have you ever dealt with over 1000 bounce-back messages? It's as if every email server on the planet uses a different bounce-back method and message. It's terribly unpredictable. And false positives are the problem I'm talking about. If you allow bounce-backs or replies with personalized messages, as a larger company, you have to let them through or you will miss something. I don't know support departments big enough to wade through this amount of junk every time you have to notify your user base of something.
    - raganwald 17y ago
      What technology can deal with spam, where the originator is actively and maliciously trying to break your filtering algorithm, but is helpless when dealing with bounce-backs, where there is no such intent?
billymeltdown 17y ago
> <em>I understand that a large email blast will generate all kinds of auto-responses, email bounce messages, and other forms of invalid replies, but with some decent filtering and a little work an organization could surely separate the wheat from the chaff. It's the least they could do to in exchange for reading their message.</em> I don't think the author appreciates how difficult this is. It's just not realistic in all situations, especially situations where the user knowingly signs up to receive an automated message. I normally use Error-To for do-not-reply, and post a valid reply-to header, but in many scenarios this isn't enough. dpcan's first comment is particularly relevant > "the email address has to resolve back to the sending server properly or the spam filters kill it"
- jodrellblank 17y ago
  I don't think the author appreciates how difficult this is. I think it's trivially easy. Hire some temp workers to do it. Heck, you could put a note "We send these emails to a lot of people and many are returned to sender unopened. If you reply to this, please type the word 'reply' in the following space to make it stand out from the returned emails and grab our attention : [ ] " Then filter those out as a priority. That could solve a lot of the problem even without a good mail filter.
  - jncraton 17y ago
    Temp workers are more expensive than CPU time. Bayesian filtering would go a long way in cutting out the auto responders, but I agree, the problem doesn't seem too difficult.
    - prawn 17y ago
      If it wasn't too difficult, big companies with loads of money would be doing it, right? Doesn't seem that many are, which suggests that the biggest and best haven't found an efficient solution they think is worth paying for.
      - mmt 17y ago
GavinB 17y ago
We have a catch-all running on the domainthat routes all e-mail sent to unknown addresses to customer service. We get e-mail sent to "no-reply" registration all the time. If we didn't do it, It's worth it not to have so many customers feel ignored the first time they try to contact us.
- eli 17y ago
  Careful with that. There are spam bots out there that just pick a domain and starting firing away many thousands of messages at <common_user_name>@yourdomain.com

43 points by numlocked 17y ago

QFT: Using a do-not-reply address to communicate ... sends a message that the organization's time is more valuable than yours. Though there are legitimate exceptions - I'm thinking, for example, of an opt-in automated status update that has the sole purpose of sending you a piece of information to which you would have no particular need to reply.

Torn 17y ago
Exactly this. There's plenty of legitimate one-way uses of email communication. When I'm getting confirmation emails from Amazon after having bought something, for example, I do not expect to be able to reply to the machine sending them out -- it doesn't make much sense. I'll go further with my argument: the ability to reply via email doesn't guarantee my communication will be dealt with in an appropriate or timely manner. Far better, I think, to go to the support section of a site and use a guided process to get my query to either answer the question myself or to have it directed down the appropriate channel.
- prodigal_erik 17y ago
  > the ability to reply via email doesn't guarantee my communication will be dealt with in an appropriate or timely manner. If they're too incompetent to handle email, I won't expect timely or appropriate action from them no matter what I do, and I'll regard doing business with them accordingly.
  - Torn 17y ago
    Having a single email address at the top of an email, necessitating parsing or eyeballing the content to understand the intentions and then forwarding it on to the appropriate people is surely more work, and less efficient, than the alternatives. It's conceptually easier to streamline support via a guided on-line process or knowledgebase type repository. Users can often answer their own questions, or if not they can at least be pointed at the appropriate department. I'd go as far as to say a large retailer would appear more incompetent if my only access to support was via a single email address.
    - numlocked 17y ago
      I completely agree if the email was the only point of access, but forcing customers onto a site is a classic case of forcing customers into a work flow that makes sense for the organization instead of providing tools to work with the customer's natural work flow.

dpcan 17y ago

False. 1) Larger companies have servers dedicated to sending out mass emails and the email address has to resolve back to the sending server properly or the spam filters kill it. 2) They need you to click the links in the email for analytics. Learning what offers get the most traction HELPS them AND YOU so they deliver relevant content in the future. 3) If you use a "support" or "sales" email as the sender and then send the mass email to 100,000 people, you are going to get 20,000 bounce-backs, littering your support email channel and support software with garbage. 4) You can't have people replying with "don't email me again" because those will get missed. Again, you can't have a large list and realistically go through all the bounce-backs looking for legitimate unsubscribe messages - the links IN the email need to be used for this. It's not fair to compare email campaigns to calling a company on the phone. Apples and Oranges.

lsb 17y ago
But these bouncebacks are predictable, no? There are several key words in them that identify them as bouncebacks, no? So make a spam filter for them, and it'll work great, especially because the mail programs aren't trying to evade spam filters with their messages.
- dpcan 17y ago
  Have you ever dealt with over 1000 bounce-back messages? It's as if every email server on the planet uses a different bounce-back method and message. It's terribly unpredictable. And false positives are the problem I'm talking about. If you allow bounce-backs or replies with personalized messages, as a larger company, you have to let them through or you will miss something. I don't know support departments big enough to wade through this amount of junk every time you have to notify your user base of something.
  - raganwald 17y ago
    What technology can deal with spam, where the originator is actively and maliciously trying to break your filtering algorithm, but is helpless when dealing with bounce-backs, where there is no such intent?

billymeltdown 17y ago

> <em>I understand that a large email blast will generate all kinds of auto-responses, email bounce messages, and other forms of invalid replies, but with some decent filtering and a little work an organization could surely separate the wheat from the chaff. It's the least they could do to in exchange for reading their message.</em> I don't think the author appreciates how difficult this is. It's just not realistic in all situations, especially situations where the user knowingly signs up to receive an automated message. I normally use Error-To for do-not-reply, and post a valid reply-to header, but in many scenarios this isn't enough. dpcan's first comment is particularly relevant > "the email address has to resolve back to the sending server properly or the spam filters kill it"

jodrellblank 17y ago
I don't think the author appreciates how difficult this is. I think it's trivially easy. Hire some temp workers to do it. Heck, you could put a note "We send these emails to a lot of people and many are returned to sender unopened. If you reply to this, please type the word 'reply' in the following space to make it stand out from the returned emails and grab our attention : [ ] " Then filter those out as a priority. That could solve a lot of the problem even without a good mail filter.
- jncraton 17y ago
  Temp workers are more expensive than CPU time. Bayesian filtering would go a long way in cutting out the auto responders, but I agree, the problem doesn't seem too difficult.
  - prawn 17y ago
    If it wasn't too difficult, big companies with loads of money would be doing it, right? Doesn't seem that many are, which suggests that the biggest and best haven't found an efficient solution they think is worth paying for.
    - mmt 17y ago

GavinB 17y ago

We have a catch-all running on the domainthat routes all e-mail sent to unknown addresses to customer service. We get e-mail sent to "no-reply" registration all the time. If we didn't do it, It's worth it not to have so many customers feel ignored the first time they try to contact us.

eli 17y ago
Careful with that. There are spam bots out there that just pick a domain and starting firing away many thousands of messages at <common_user_name>@yourdomain.com