By Chris | November 12, 2008
I just spent the last 30 minutes deleting over 2500 blog comments “Awaiting Moderation” because my filters indicated they might be spam. All of them were. I was deleting them 25 at a time through the WordPress interface, before a Google search showed me how I could delete them all in one fell swoop using MySQL (the database on my server that holds all my posts and comments). Isn’t the free transfer of information great, and frustrating at the same time? I also installed some software that should delete more spam comments before I even have to look at them.
I was prompted to address my spam problem after reading an article about the Economics of Spam (HT: Al Roth’s new blog) . It summarizes the research of UC Berkley computers scientists who hijacked a spam network and used it to test the conversion rates of unsolicited electronic advertisements. An excerpt:
While running their spam campaigns the researchers sent about 469 million junk e-mail messages. The vast majority of these were for the fake pharmacy campaign.
“After 26 days, and almost 350 million e-mail messages, only 28 sales resulted,” wrote the researchers.
The response rate for this campaign was less than 0.00001%. This is far below the average of 2.15% reported by legitimate direct mail organizations.
“Taken together, these conversions would have resulted in revenues of $2,731.88—a bit over $100 a day for the measurement period,” said the researchers.
Scaling this up to the full Storm network the researchers estimate that the controllers of the vast system are netting about $7,000 (£4,430) a day or more than $2m (£1.28m) per year.
It amazed me how little spammers make from their campaigns. I see two implications from the study:
- If you ever purchase products you hear about from Spam, you are imposing a large negative externality on the rest of country. In this instance, 350 million emails resulted in only 28 sales. In other words, for every sale, 12.5 million emails don’t convert. Spammers will continue to send out unsolicited emails as long as the marginal cost of an advertisement equals the marginal revenue it brings in. The marginal cost of an email is so low for the spammer that even one likely sale provides the incentive for millions of emails. Purchasing a product introduced to you by spam is a lot like giving money to beggars, except there is nothing morally redeeming about it.
- We could eliminate spam if we charged .1 cents per email. 1/10th of a penny would do it if the conversion rates in this study are representative of the industry as a whole. I think that we could integrate payment into email without reinventing the wheel either. People could simply buy credits from their email provider (gmail, yahoo, or some other third-party). If the email they sent was accepted, there would be no charge. If it was not marked as legitimate, the sender would be charged. People who didn’t guarantee their email with payment from a third-party would still be able to send email, it would just be more likely to be filtered out. I imagine most people wouldn’t mind paying a few cents a day to make sure that relevant emails to people they may be emailing for the first time would’t end up in the spam filter. I know I wouldn’t. While this idea has been proposed before, the study above seems to show just how cheap it would be to implement.