So this has been a topic before, but I don't think has been given the right light it needs to have in today's ever changing landscape of the good guys (ISP's) trying to stop or identify the bad guys (spammers and bad marketers) or even know who the good senders are without taking their important efforts off finding the phishers out there

As some of you know, there has been a significant and fundamental shift in how receivers perform anti-spam functions. In the past and still a little today, receivers would review content mostly for triggers like bad words, porn and drug connotations, and other such moral and or offensive issues that weren’t so accepted years ago. As time has moved on, we have noticed a more accepting audience to some of the above-mentioned content (i.e. someone needs Viagra…no for real) and the false positive rate for content filtering got worse. So what are we to do?

Well today most receivers look to filter out email based on your behaviors today or on what we call reputation. Think of it as a credit score for your email program. You’ll know where you stand, how email receivers are evaluating you, and what you need to change about your program to improve your delivery rates. Things such as complaints, unknown users rates, spamtraps, and volume play a role in that. Think I am making this up? My good friend George Bilbrey said it perfectly here in this post “When it comes to email blocking and filtering, reputation -- not content -- can be blamed 83% of the time.”

So your saying to yourself either “I know this Dennis” “or what does this have to do with the subject line?” Here are few thoughts first:

1. Each mode of messaging you send today or class of email (marketing vs. transactional vs. mixed mode vs. alerts) has a different importance today to those who receive it. How long they look at it? How easy or fast they might unsubscribe from it? or even complain about it (heaven forbid)
2. Each sender or entity also has their own set of problems or criteria that drive or hurt their email programs. Some buy lists. Some do not perform confirmation opt it. Some make up an email strictly of images only with no HTML (images are blocked by default within most email clients). Some use lists that are five (5) years old and haven’t been used since.
3. Just like today, your social security or driver’s licenses identification belongs to you and your habits, why would you want someone to share that identification? Or ruin yours? Think of it like the black sheep (not that I have one here) in your family… they can ruin the name of the family so easily without even caring about your real identity or habits as the white sheep.

Today I see so many people sending mixed messages through the same IP without even blinking an eye at understanding or concept of segmentation or risk of their email stream if their marketing messages are getting their transactional email blocked. Yes believe it, marketing message do get blocked or complained about more than transactional billing information (unless that bill is overdue) – seriously, a pay day loan company here in Dallas/Fort-Worth sees their overdue transactional notices marked as spam more than their marketing messages!

Today I see so many ESP’s who still allow customers to share IP space. That means if the guy sharing the same IP decides to send to one of those 30 million-email addresses on a CD’s he bought online, then you could get blocked as well for his mistake or habits.

Folks, don't put all your eggs in one basket is what I am trying to say here. Segment your email streams.

Segment by:

1) Customers or entity.
    a) Give them their own identity. If they mess up or someone else does, then you can mitigate the problem to a smaller space of IP’s. *HINT* This also helps the receivers from having to block large ranges of IP’s!

2) Class of email.
    a) Marketing vs. transactional vs. mixed mode vs. alerts. If your marketing messages get blocked, then your transactional or mixed-mode messages won’t since they run off a different IP set

There are other ways as well to segment. Product lines purchased, activity of users, geography, purchased something, visits to your website and the list goes on and on.

Talk to your marketing or IT folk’s people. Make sure you start to assign separate IP segments to your customers or yourselves. When you separate customers, your divisions, and or classes of email... you can better maintain and monitor the reputations associated with them to see who is causing your problems. Maybe even determine which customers you need to fire (yes I said fire a customer!). Don’t put all your eggs into one basket.

Take a free look at your reputation today using ReturnPath’s reputation tool at http://www.senderscore.org. It looks at single IP and or domains as well. Funny? Isn’t that monitoring each separately as it should be measured?

-Dennis
Eloqua