A New Kind Of Anti-Spam?

Here's a new anti-spam service which takes a somewhat different approach.

RI-based Mail Cruncher works, not by looking at content, but by rating emails according to the reliability of their sender. "In business, I decide to trust other companies based on how long they have been around, their location, who else does business with them, and their record of reliability," a press release from the company quotes April Lorenzen, creator of the Mail Cruncher service, as saying. "We just applied the same common sense to sorting email. Our customers say it works well for them, saving time and aggravation every day."

The Mail Cruncher email sorting service uses Outbound Index ratings exclusively to sort email. Messages with a high rating go immediately to the subscriber's inbox. Messages with a low rating are held back. Once a day, Mail Cruncher subscribers are sent an email with a sorted, color-coded list of suspicious emails that can be scanned in seconds.

The ratings are based "principally on statistical facts such as domain age, relationships between server and domain, and sender stability". There's no attempt to run Bayesian filters or other approaches to measure the spamminess of an email. It's done entirely by looking at the 'from' email address.

The Mail Cruncher list also groups domains, "so if a subscriber receives 17 messages from the same domain, the messages appear together for faster skimming". It also displays the sender's user name, such as "nwyiyvq," is displayed, not just the often-misleading name ("Victoria") shown by most email inboxes. Finally, a subscriber can read the text of a suspicious message safely within the Mail Cruncher environment "without triggering webbugs or attached viruses, without displaying any objectionable images that might be in spam, and without the sender knowing the email was opened".

An actual Mail Cruncher list with the above features is demoed here. The Mail Cruncher service costs $3 per mailbox per month.

Comments

[April Lorenzen]

Very nice mention, thank you.

To clarify, the Outbound Index uses the IP address of the sending server, as well as the domain part of the envelope-from, in cross-referencing facts about stability, relationship, and longevity.

We do not use the "header from" - it's often not related to the sending server, such as in mailing lists. The list "envelope-from" might be lists@example.com, and the "header-from" would be the individual writing to the group, such as JoeDoe@aol.com. The Outbound Index in this case would use "example.com", paired with the IP address of the sending server, in evaluating how to classify the message.

Thanks,

- April Lorenzen