Baby and Bathwater Redux

Katie Hafner’s article in the New York Times Circuits section today about spam and false positives is right on the mark. Spam filters are still evolving, and spammers are evolving right with them. Although the flood of spam is largely stemmed by a good filtering app, the results for consumers are still spotty: false negatives are irritating, false positives can be very painful (as the article suggests), and the process still consumes a little too much time. While the article nails the consumer problem, it does miss the corresponding business problem around false positives (see below).

But things are getting better. While I wrote generally about how email is here to stay a couple weeks ago, there are a couple other things I’d point out after reading Katie’s article that are making the email landscape a brighter place of late:

First, even better than the Bayesian style filters referenced in the article are community-based filters. The leading one is run by Cloudmark and is called SpamNet. SpamNet relies on a community of 1 million hardcore email users voting on whether email is spam or not. I’ve used SpamNet for over a year now, and while it’s not perfect, it’s pretty good at reducing both false positives and false negatives to a tolerable level, and it’s very easy to use (but only with Outlook for now).

Second, a few companies in the email industry — Ironport (Bonded Sender) and Return Path included — are hard at work on solutions to the false positive problem that won’t leave false negatives behind. Once these solutions reach maturity (still 6-12 months away), I think consumers will notice a quantum leap improvement in managing their inboxes.

Finally, one thing I’m always trying to encourage people to realize is that this problem is not just an annoyance for them personally…but it’s an annoyance to legitimate businesses everywhere. Businesses who uses email to reach their customers — when customers request the email — are consistently finding that anywhere from 5 to 50% of their emails are blocked or filtered (with an average of 19%, according to our research). Talk about an ROI buzz kill and a CRM nightmare!

So hang onto those babies out there, consumers and marketers alike…the bathwater really will go down the drain soon!