Spam filtering is an extremely important function for Web-based email providers, as it often decides which provider users should go with. Microsoft (News - Alert) recently posted spam statistics comparing its Hotmail service to a few of the most popular options, including Google's (News - Alert) Gmail and Yahoo Mail, and unsurprisingly declared Hotmail the best at keeping spam out of users' inboxes – with Gmail in close second. Obviously, Google had something to say about this.
Google stepped forward today, and declared its Gmail service the leader in anti-spam, citing its own statistics which indicate less than one percent of email in the average Gmail inbox is spam, while its rate of "false positives," legitimate messages accidentally marked as spam, is also less than one percent.
If the statistics of both companies are true, this puts Google's anti-spam system head and shoulders above Microsoft's, which reported in February less than three percent of the average Hotmail inbox is spam, with no false positive rate provided.
Google says its system can provide better results due to a more stringent definition of spam than Hotmail's. While Microsoft's system measures only unsolicited mail, Gmail includes both unsolicited and solicited mail which users consider to be spam, also known as "gray mail."
The user-defined system is what Google says allows it to provide such accuracy in its anti-spam, letting Google consistently update and fine-tune its spam definitions based on which messages users manually mark as spam.
Even messages users consider legitimate are evaluated. The effectiveness of Gmail's anti-spam system is no surprise, since Gmail was created back in 2004 in part to prove Google could eliminate spam better than other services.
Since going public in 2005, Gmail's user base has grown steadily, providing Google with more feedback and resulting in better spam protection. The war on spam, however, continues as Google focuses on keeping false positives to a minimum since missing an important message is a major concern.