Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

I'm currently doing research on a better way for emailing customers our news letter and not using an ESP such as vertical response since most users there use purchase list and ruin the shared IPs for everyone.

My question is, do you really need to be sending hundreds of thousands of emails from your mail server IP address to be put on a "accepted" list for ISPs? I've read you need to be sending a constant amount of emails from your server. I don't understand how this can even be true since no company in the world sends that many emails to a specific ISP.

Also do most email companies such as: yahoo, google and random ISP email providers (such as time warner, cox, etc) simply use DNSBLs to check each incoming mail to see if it's spam?

Thanks for all your help!

share|improve this question

5 Answers 5

up vote 4 down vote accepted

Different mail servers and different organizations use a variety of different ways to combat spam.

  • DNSBL
  • Various email verification techniques.
  • Various mail server verification techniques (ie. reverse lookup for PTR record).
  • GeoIP filtering
  • Email message content analysis
  • Checking the number of "bad addresses" you are trying to send to
  • Checking how many emails you have sent to their system
  • Checking how many connections you have established with them in the last __ minutes, etc.

The list goes on and on...

The short answer to your question is, "No" you do not need to be sending a constant (or even large) amount of mail to any system to "prove" you're legit.

If you're sending legit email then the best thing you can do is monitor who is blocking you, find out why, and work with them (one on one) to get them to put you on their "accepted senders" list or "whitelist" or whatever they want to call it.

On some systems this is an automated thing... on others it is a human thing. I believe I read somewhere that someone even had an idea of putting up a bond of some sort to prove you weren't a spammer (ie. they catch you spamming you lose the $$$).

If your email list is truly legit there is no reason why you can't use your own mail server and program such as PHPList (or similar products) to send out your newsletters/email, etc.

TALK TO YOUR ISP AHEAD OF TIME and tell them what you're planning on doing.

Even thought your email may be CAN-SPAM compliant, most ISPs have their own unsolicited (or bulk) email policy.

You'll find that most ISPs are very friendly if you talk to them AHEAD OF TIME and tell them your plans. Many will even forward on the "complaints" to you so you can put the complainer on an "opt out" list.

share|improve this answer
    
Okay, so each ESP has their own verification procedures and techniques and some also outsource this to other companies that have DNSBLs, etc? –  kernelPanic Jun 30 '09 at 4:43
    
Thank you so much, you answered my questions! You rock!! =) –  kernelPanic Jun 30 '09 at 4:50

Almost no ISPs offer a "whitelist" service. in fact, I don't think any offer one any more. Just make your mailouts as friendly as possible to minimise the chance of unwanted blocking. For example:

  • Sign all your outgoing mail with DKIM
  • Set up SPF records for your mail domain
  • Use a separate subdomain for bulk mailouts
  • Sign up to feedback loops with major ISP targets of yours (eg AOL, Windows Live, etc) and unsubscribe people who mark your mail as spam
  • Remove non-working email addresses from your lists ASAP
share|improve this answer
    
Different ISPs call it different things. While the term "whitelist" might be a bit dated the concept is very much alive. Take Yahoo's "Mail Bulk Sender Form" as just one example of many: help.yahoo.com/l/us/yahoo/mail/postmaster/bulkv2.html –  KPWINC Jun 30 '09 at 5:13
    
Don't bother with SPF, it's used as a sign of spam; messages with valid SPF are actually more likely to be spam than those without. –  MarkR Jun 30 '09 at 6:53
1  
I respectfully disagree. SPF isn't directly about spam (or even anti-spam. In our experience SPF (if implemented PROPERLY... and that's KEY) will improve the delivery rates of your email. Your mileage may vary, I'm just sharing with you what I have personally experienced. SPF implemented incorrectly will be your worst nightmare (been down that road too). –  KPWINC Jun 30 '09 at 15:49

I won't duplicate what KPWINC has said, as that is pretty much what I would have written. The only thing I would add is to give very careful consideration to to body of the messages in question in order to reduce the risk of being flagged as spam. If possible send a test message through a spam filter and see what score it returns.

share|improve this answer
1  
One can also Google for terms such as "spam score" and "spam analysis" to find various tools and services that you can test your messages against. Like anything else, scoring systems vary... but if you know nothing about spam scoring, etc its worth it to try a few of these services. Here is one for example: politemail.com/spamscore –  KPWINC Jun 30 '09 at 5:19

I've never heard of such a requirement, and I've maintained a newsletter package (phpList) for a small company for a few years.

In my experience, the larger email providers seem to have their own unspecified method of filtering spam; the smaller ones will often use a simple anti-spam package (or simply the MTA) to test messages against a DNSBL.

Many places (Yahoo comes to mind) are also using greylisting, so you might see some delay in getting your messages sent out.

Make sure your lists are clean (the recipients actually want your newsletter). This will reduce the likelihood that the recipient will tag your mail as spam and get your IP banned from the entire company (if enough people do likewise). Also, take steps to make sure your messages aren't "spammy" (using a generic "From" address such as "webmaster" or having more HTML markup than actual text). This will lower the chances that automated spam filters will flag your message.

One final thing to consider: if you are shopping for hosting, tell them that you plan to do a legit, verified opt-in mail list (which you should be) and ask them to pre-screen the IP they will be assigning your host against the major blacklists to make sure the IP hasn't been tainted by a prior spamming customer. Once, after upgrading to a new server (at the same hosting provider) I found that a good portion of the newsletter emails would be rejected and getting removed from the banned list at large places was frustrating (and often temporary). I eventually got them to give me a new block of IP addresses, since even the hosting provider couldn't get some of the larger places to remove my addresses.

share|improve this answer

No, we use an IP-address based reputation service which is internally managed automatically and contains very few white/black list entries (except for our customers' own addresses) and use that.

Nobody needs to send thousands of clean messages to get on to the "good senders" list, just a single clean email will suffice to get you ahead of anyone who we've not heard from before.

But the trick is to not get your messages classified as spam in the first place. Techniques listed in other responses are helpful, I'll summarise/ add my own:

  • Don't share a relay IP address with spammers (this is really the main one)
  • Don't put URLs in the message whose servers get compromised by malware
  • Do set up a sensible believable and forward-resolvable reverse DNS for your relays (this is pretty much mandatory for most recipients nowadays)
  • Don't obfuscate your messages in any way
  • Don't use "microdot" images in your messages.
  • Don't put any strange HTML in which is likely to look the least bit malicious, for example, don't put scripts in (Mail user agents don't execute them anyway), or use Flash, Java, etc which
share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.