Sometime in 2004 or 2005, the world of e-mail will likely experience a dramatic overturning of e-mail marketing strategies and infrastructure. Implications for online retailers are enormous. The catalyst? Spam, of course.
Thanks to spammers, users’ inboxes are highly polluted with irrelevant commercial messages. The problem has grown to the point where there is widespread agreement that something must be done to stop it-even something as radical as changing the very rules governing the sending and receiving of e-mail.
As we all know by now, legislation hasn’t stopped spam. Anti-spam software doesn’t help much either, since spammers can outsmart software algorithms with random word “poetry.” One of the most viable solutions now on the drawing board is a technical one that’s been dubbed the “Penny Black solution” or “puzzle solution” and is backed by Microsoft. Bill Gates has promised to end spam by 2006, and this puzzle solution seems likely to be the key to delivering on that promise.
What is it?
This article explains how this puzzle solution will impact your business. You’ll learn how it works, what it means in terms of your interaction with customers and what you’ll need to do to be puzzle solution-compliant by 2006. The implications to Internet retailers are potentially very large.
The only thing that prevents your physical mailbox from being flooded with thousands of offers every day is the fact that direct mailers have to spend money on postage. If postage were free, you’d be inundated with so much physical mail you’d have trouble finding your mailbox, much less the mail you really want.
That’s what’s happening with e-mail: spammers pay practically nothing to send e-mail, so there’s every incentive to keep sending as much as possible in the hopes that somebody will buy something. The puzzle solution seeks to impose an indirect cost (“friction”) on sending e-mail. If e-mail costs even a penny to send, spammers are instantly out of business since they can’t recoup even one penny per e-mail by indiscriminately spamming people with irrelevant offers.
There’s no real penny involved here, though. The puzzle solution adds a cost to sending e-mail by requiring an expenditure of CPU cycles, not money. Under this solution, a mail server (the machine in your company that sends your e-mail) would have to spend approximately 10 seconds of CPU time to solve an encryption puzzle and stamp each outbound e-mail. It is this 10 seconds of CPU time that results in adding costs to the sending of e-mail, and it is precisely this friction that stops spammers cold.
If the puzzle solution is adopted by the Internet community, it will likely be based on the mathematical algorithms already devised by Microsoft engineers, which will probably be distributed to the software developer community free of charge. These algorithms will require sending mail servers to execute a routine of computational gymnastics for each e-mail waiting to be sent. Because the computed solution is unique to each e-mail, mail servers cannot run these calculations ahead of time and, for example, queue up a stack of solved puzzles. Rather, they must process each e-mail as it is about to be sent. The end result is a significant slowing of outbound mail servers.
On the receiving side, ISPs and other organizations are free to decide whether they wish to pay attention to puzzle solution stamps. There is a tiny cost associated with analyzing incoming e-mails to verify their puzzle solution status, but this cost is marginal, perhaps a few thousandths of a second in CPU time, and is vastly offset by the benefit of rejecting e-mails that aren’t stamped with the puzzle solution.
Because of the simplicity of checking for puzzle solution compliance and the tremendous cost benefit of avoiding spam, ISPs and organizations have every incentive to filter incoming e-mail based on puzzle solution status. Activating this filter will be no more complicated than the IT team downloading and installing the latest “puzzle solution compliant” update of their mail server software.
The puzzle solution will be transparent to end users, e-mail marketers, and everyone but the IT personnel. End users who send and receive very low volumes of e-mail on a daily basis will notice no difference. Marketers sending e-mail newsletters or updates to small lists (up to a thousand recipients) won’t notice much difference either. But organizations that engage in high-volume e-mail activities will notice a big difference indeed: the sending of e-mail will suddenly become more costly in time and resources.
The cost of compliance
Today, a company can easily send 100,000 e-mails per hour using a single PC running mail server software. The cost of the server is only a few thousand dollars, and there’s no incremental cost for each e-mail.
Under the puzzle solution, however, a single mail server can send only one e-mail every 10 seconds. That comes out to 360 e-mails per hour, or around 8,600 in a day. A list of 100,000 e-mails would require 11 days to send.
That’s the difference: one hour today vs. eleven days under the puzzle solution.
So, without any infrastructure changes other than updating the mail server software to be puzzle-solution compliant, a retailer could send 8,600 e-mails per day. If a marketer wants to send more e-mail every hour, there’s only one way to do it: buy more machines.
With a farm of 10 e-mail servers, you could send 86,000 e-mails per day. Every e-mail would be stamped “puzzle solution compliant,” and all e-mails would be accepted by the receiving servers.
This addition of servers is, of course, expensive. And that’s precisely the point of the puzzle solution: by making the sending of a large volume of e-mail expensive, spammers are stopped cold. Not only that, but legitimate e-mail marketers will get more discriminatory in sending e-mail. All those free daily e-mail newsletters will vanish and be replaced by weekly or monthly e-mail newsletters. Suddenly, sending a daily e-mail to 100,000 subscribers isn’t free.
0.5-1 cent per e-mail