A study on messaging and Web security we conducted earlier this year asked messaging-oriented decision makers about changes in their spam filter false-positive ratios over time. While we found that one-third of midsized and large organizations reported their false positive ratios were improving, 53% said they were staying the same and 14% reported they were actually getting worse.

An analysis by Opus One found that some spam-filtering solutions can generate false positive rates as high as 0.85%, or one in every 118 e-mails. Given that the typical user in a large company receives 102 e-mails each workday, or about 26,500 e-mails each year, that translates to as many as 225 lost e-mails every year. The problem is compounded by the fact that not all employers allow their employees to review their spam quarantines – even among those that do, not all users check their quarantine for false positives or do so on a timely basis.

What if a falsely identified e-mail contains an order, an inquiry from a prospect or an important communication from a client? That could cost you significantly, both financially and in intangible ways, such as offending a key prospect by ignoring his or her e-mail to you.

The problem with false positives could become worse in a unified communications and unified messaging environment that results in more content being dumped into your mailbox, making it more difficult and time-consuming for users to pore through quarantined content. It could become worse with a data leak prevention (DLP) system that looks for sensitive or offensive outbound content too aggressively and blocks legitimate messages from leaving the organization on time. As our research shows, it actually is getting worse for one in seven organizations in North America.

Michael Osterman is principal analyst of Osterman Research.