Email Spam Filtering
Spam doesn't just distract your employees; it can also expose your systems to malware and phishing attacks.
We estimate that over 95% of all email traffic is spam. Businesses lose millions of dollars in productivity and additional infrastructure expenses due to spam. To make matters worse, organizations that maintain in-house email servers are fighting a losing battle, because spammer tactics constantly change. To defeat spam, you must stay one step ahead of the evolving threat.
Multi-layered Spam Filter
At Rackspace, we continuously update the spam filtering system we use for our Rackspace Email boxes in a multi-layered process that eliminates 98% of spam—with nearly zero false positives. You get spam filtering at no additional cost with all Rackspace Email accounts.
Layer 1: The Gateway Scan
As soon as an email arrives, our gateway servers try to match the sending IP address to an aggregated blacklist compiled from multiple spammer tracking systems. The servers analyze the email message in comparison to other arriving mail. If a large number of emails arrive simultaneously from a single IP, or are addressed to users that do not exist in our system, it could signify a spam attack, and the servers block the offending email. If the sending address is from a domain in our system but the mailbox does not exist, the servers block the email.
Layer 2: Cloudmark® Scan
We scan all email with Cloudmark's industry-leading spam detection software. Cloudmark uses Advanced Message Fingerprinting™ to detect viruses, spam, and phishing. Advanced Message Fingerprinting uses algorithms that detect spam across all languages and character formats. These algorithms update every 60 seconds based on worldwide feedback loops and the latest spam tactics.
Layer 3: The Message Sniffer Scan
We scan email with Message Sniffer from ARM Research Labs. Message Sniffer relies on pattern recognition and machine learning technology to detect spam and malware. It searches the entire message for spam and malware features, including unusual headers, message source behaviors, structural artifacts, obfuscation techniques, binary and image signatures, email and URL targets, unusual code fragments, and even coding styles.