Spam.

If you’re on the Internet, odds are you know what it is. How could you help but? Unless you keep your email address in a safe and never give it out, some spammer has gotten ahold of it at one time or another—which means they all have it. If you’re like me, that means your mailbox is filled to overflowing with people who want to enlarge both your penis and your breasts, sell you “V1agra” or “C1alis” or software or printer supplies or mortgages on the cheap, get you to help them get their money out of Nigeria, have you send them your or just fill up your box with missives in languages you cannot read.

I was fortunate enough to come to the Internet back in 1992, when the “August” page had not yet been torn off of the Internet calendar. I enjoyed my first couple of years of email entirely spam-free. I still remember the furor caused by Canter & Siegel’s legendary First-Ever USENET Spam that heralded the start of the Internet spamming industry (and was just one more element of the “eternal September” that’s been going on ever since). If you’re like me, you wish you could find a time machine and go back and stop them…but even if you did, you know someone else would come up with the idea sooner or later.

What percentage of all Internet email messages do you think spam constitutes? 20%? 30%? Try 65% of all email messages sent over the Internet, according to Symantec. That’s right—out of every 20 emails sent, 13 are spam. Multiply that by the billions of emails that are sent daily…what a nightmare. It’s the tragedy of the commons on a grand scale.

Think of how much more efficiently the net would run if we could flip a switch and make all that spam go away. Think of how much faster files would download and “legitimate” mail would be sent.

I doubt you will find more than the occasional oddball here or there who is unwilling to admit that spam is a problem. I don’t think anybody normal likes the stuff…I doubt there’s one of my readers who would be sorry to see it go. This kind of begs the question…if the average person doesn’t actually want the stuff, then why do the spammers keep churning it out? You’d think they’d have gotten the hint by now that this is not an optimal way to sell a product.

But the thing is, spam may very well be an optimal way to sell a product, as this quote from a Salon.com article indicates:

According to a former [Davis Wolfgang] Hawke associate, the […] spammer boasts of earning “six figures” and often carries around a wad of hundred-dollar bills in his pocket, totaling thousands of dollars.

What’s going on here? How can someone be making so much money selling something almost nobody wants?

The key word here is almost.

In a way, it goes back to an old ethnic joke wherein the person of the ethnicity, nationality, or hair color of your choice is standing next to a watermelon cart with a sign reading “Watermelons, $1,000,000 each.” A passer-by asks, “Are you crazy? There’s no way you can make a living selling watermelons for a million dollars each!” The watermelon-seller replies, “I only have to sell one.”

In similar vein, spamming is so incredibly cheap to do that the spammer only needs to sell a handful of his product to make a profit. When you think about it, with today’s high-speed Internet access and software programs specifically designed to automate the spamming process, sending an email costs almost nothing at all. You can send out thousands of emails per second if your connection is fast enough. How much does one thousandth of a second of electricity cost? Of wages? If only one in one thousand people to receive a spam bought, and it was sent to a million people, that’s a thousand sales right there.

And the less moral kinds of spam enjoy an even greater level of success than that—because many people are depressingly gullible. A recent study has revealed that users are unable to tell fake “phishing” (fraudulent social-engineering for the purpose of stealing your credit card number, identity, etc.) emails from genuine business emails. These are the messages that claim to come from PayPal or Visa or eBay insisting that they need your credit card information, but really point to “PayPaI” or use HTML redirect tricks to take you to a non-Visa site. (Even more depressing…I took their test yourself quiz and only scored 60% accurate myself—though I still maintain they cheated by removing the actual URLs to which the email HTML pointed and leaving just the ones the email claimed they pointed to; in a “real-world” situation comparing the claimed to the real URL is one of the main methods I use to tell the difference.)

At any rate, spam isn’t going to go away soon. Proposed legislative solutions lack teeth; most technical solutions are either unrealistic (“let’s charge everyone for email!”) or still in the planning stages, and what’s needed is a solution for here and now. Fortunately, there is a current technical solution that seems to work pretty well as a stop-gap.

I’m not going to duplicate the LiveJournal post in which I describe exactly what it is and how I set it up, but suffice it to say that, at this point, Bayesian filtering seems to be the best answer to combatting spam on a personal (or server) level. There are software systems that use Bayesian filtering available for all OS platforms (it’s even included in OS X’s default email application) and it works remarkably well. I’ve seen the spam in my inbox fall from dozens a day to four or five per day at most, and even that will fall off as the Bayesian filter learns more about what spam looks like. My mailbox is down from fully half spam by subject-line to readable again. I no longer miss messages because they’re buried in spam.

Maybe someday the Internet calendar will flip on to October and the spam will stop. It’s something to hope for. Until then, we just have to filter and make the best of it.

Advertisements