Smart e-mail delivery
The bottom line is these ISPs continue to deal with an increasing amount of spam each day and with that they continue to place new walls in front of their e-mail systems to combat the problem. A recent security study by Cisco estimates that spam volume has doubled in 2008 from 2007 with nearly 200 billion spam e-mail messages sent each day. The more crafty spammers get in sending out their mail the more scrutiny each ISP places the e-mail sender under. So how are we as an industry to stay ahead of the curve or at least stay with the curve?
This is not a new problem. The idea of incorporating ISP specific rules under which an SMTP server would throttle email to specific domains has been incorporated into most of the enterprise class mail transfer agents (MTAs) on the market. The idea behind these rules is that the MTA vendor determines the parameters by which a specific ISP prefers to receive bulk e-mail. These rules are then used as the basis for sending email by the MTA during deliver to the ISPs.
As the vendor determines a rule has changed the rule is updated in the server. Incorporating these intelligent delivery rules into the MTA has helped over the last several years protect a senders IP reputation and kept senders from unnecessarily raising red flags with ISPs. These pre-canned rules are a good baseline, but often, one set of rules does not apply to all senders. Several factors, including IP reputation, e-mail content, "from" addresses and frequency, can affect the way a sender is viewed by an ISP. Having a solution that is flexible enough to allow for these variations becomes important to good deliverability rates.
Just as MTA vendors began incorporating domain level delivery rules into their platforms, the next generation of MTAs will take this concept to the next level. The future of intelligent e-mail delivery lies not only in predefined rules but also in real time adjustment of delivery rates based upon the feedback the MTA is receiving from the ISP during e-mail delivery.
Some of this intelligent analysis is already happening with some MTAs in their ability to determine between grey listing deferrals and ordinary deferrals. The next generation of MTAs will build upon this intelligent analysis of bounce messages coming back from the ISP and determine the proper course of action in real time. This could mean slowing down delivery rates, dropping and re-establishing connections or a host of other possibilities.
As the ISPs get smarter about combating spam, so do our e-mail delivery servers need to get smarter about the way legitimate e-mail is delivered.