Get Ready for Class Discrimination

All email looks alike to your computer. It can’t tell whether your mother is asking for help understanding her health plan; a colleague wants information about a marketing promotion you’re working on; or a company is sending a purchase receipt, newsletter, or local store sales notice.

As you know, it can’t even tell the incoming email is a stupid attempt to sell an herbal extract promised to expand a body part you don’t even have. The lack of ability to discriminate between different classes of email is a problem exploited by spammers.

Who should determine what email makes it to your inbox? The problem is you can’t! Conspiracy theorists might find fertile ground proving the proverbial “they” are at work, redirecting or even deleting those important email messages from your mother and colleague while flooding inboxes with useless clutter. As the sock monster appears to devour a sock or two every time we do the laundry, the bit monster is constantly swallowing up our email. Some soon-to-be-published consumer research indicates 67 percent of recipients have lost messages to spam filters.

E-mail gateways, from those operated by colleges and small businesses to large corporations and big ISPs, struggle to determine what kind of email hits their servers to handle it appropriately. E-mail from friends, family, colleagues, business associates, prospective customers, partners, the local little league, companies we do business with, and so on are important. We must be able to rely on them getting through to us. There’s no way today to guarantee “they” aren’t deleting or redirecting messages we depend on and anticipate.

Who controls what email we get and what ends up in the bit bucket? E-mail is filtered by a host of different technologies and processes. Blacklists (also called block lists) and whitelists attempt to identify the email source (normally by looking at the IP address of the sending server) to determine whether to let messages through.

Filters on email servers and the end user’s computer guess whether an email is spam as email hits the server and the inbox, respectively. Both attempt to pattern the email in more or less sophisticated ways to see if it somehow resembles spam.

Probe networks consist of dummy email addresses or addresses used only to register with a controlled set of sources. If these addresses receive messages other than those signed up for, they have, by definition, been compromised. The email is unsolicited and, therefore, spam.

I don’t care how smart those guys writing anti-spam software are. If guessing is involved when determining whether an incoming email is spam, mistakes will be made. Mistakes, or false positives, make email less reliable.

We need email classes. What if my computer could discriminate between different classes of email and treat each differently? It could alert me about incoming personal message on my wireless or handheld device. My purchase receipts could always ended up in my designated purchase receipt folder.

Wouldn’t it be great if my trusted newsletters simply showed up with a marker indicating they are, in fact, from trusted sources? I may want some unsolicited bulk email as well. In that case, stick it all in the bulk folder. I’ll check it on my own time. If corporations could say you’re on their list of approved vendors, your email would always get through to the intended recipients.

These are not fantastic scenarios. All email could be classified with some updates to the email infrastructure. Once sender identity is established in a more secure manner than it is today, which will happen over the next few years, it becomes possible to trust assertions made in the email about its class and origin. With access to these assertions, a receiving computer can begin making smart decisions about how to handle messages.

We must be careful this doesn’t evolve into a mechanism for describing message content. Content descriptors would have all kinds of free-speech ramifications. What I suggesting is no different from what FedEx and the USPS do with atom-based email today. E-mail should be classified (maybe with a variable delivery fee) depending on its type and source.

Get ready for a world where your email will be required to be classified and be discriminated against unless it adds value. Once a receiving computer can securely establish who a sender is and what class that sender’s messages belongs to, it has all it needs to determine what to do with every piece of email, long before a recipient sees it. The individual will have greater control. Senders will be held accountable for the email they send. Add value and your messages are anticipated. Don’t and the bit monster will swallow them, guaranteed.