- The Washington Times - Sunday, February 22, 2004

Many companies forgo paying for filters to block unwanted e-mail, fearing that legitimate messages will be blocked.

Worry about these “false positives” — wanted e-mails incorrectly labeled as spam — is the main reason companies are not buying spam-filtering products.

“It probably is a necessary evil,” said Andrew Lochart, director of product marketing for Postini, a Redwood City, Calif., e-mail-security company. “I don’t think we’re ever going to reach that Nirvana of stopping 100 percent of spam with no false positives.”

Some 55 percent of companies without spam filters fear false positives, according to Radicati Group, a technology consulting firm based in Palo Alto, Calif. And 75 percent of companies with filters said preventing false positives was more important than stopping spam.

No spam-filtering product guarantees 100 percent accuracy. Radicati says corporate e-mail users lose between one and five legitimate messages each month. The company suggests the loss of one important message could be far more inconvenient than a mountain of spam.

ReturnPath, another Silicon Valley consulting firm, said up to 17 percent of legitimate messages are blocked by spam filters used by Internet service providers.

The loss of legitimate e-mail cost U.S. businesses about $3.5 billion in 2003, according to San Francisco-based Ferris Research.

Honest marketers are among the biggest victims of false positives. Permission-based marketers — those sending e-mail ads only to people who asked for them — lost $230 million in 2003 due to messages being incorrectly blocked, Jupiter Research says.

Spam-filtering companies insist their products are more than 99 percent accurate. But perfection, Postini’s Mr. Lochart and others say, is unlikely.

Many filters can be trained to detect spam over time. But these systems are not perfect because many e-mail users have accidentally labeled legitimate messages as spam.

To prevent false positives, some e-mail users set their spam filters to be less stringent. But this lets more spam through, analysts said.

“If you scale back the filters to eliminate the false positives, then why have the filters to begin with?” said Clarence Morey, product marketing manager with NetIQ, a San Jose, Calif., firm that produces the MailMarshal spam filter.

E-mail security companies disagree as to what filters should do with spam when it is captured.

Most put spam in a “quarantine” folder, which users can access to determine whether legitimate messages were incorrectly labeled. But critics argue that e-mail users should not be asked to look at a “spam” folder, since they purchased a filter to avoid spam altogether.

The imperfection of spam filters has fueled a push toward a system that would allow messages from certain “trusted senders” to be immune from spam-filtering devices. Marketers have pushed a system called “Project Lumos” that would require senders of e-mail to verify their identities.

And some Internet providers are testing systems designed to filter out messages that come from anyone looking to hide his identity. Most spammers employ a series of computer tricks to mask the source of their messages and avoid accountability.

Sign up for Daily Newsletters

Manage Newsletters

Copyright © 2021 The Washington Times, LLC. Click here for reprint permission.

Please read our comment policy before commenting.


Click to Read More and View Comments

Click to Hide