Thursday, April 23, 2009

E-Mail Notifications: An Outline for a New E-Mail Standard

If people fear e-mail, it is because it is a source of important information about things to do — and it is deeply flawed. Important messages don’t get through, and simple notifications have to be dressed up as letters, with about a page of superfluous words added, to improve their chances of getting through. At the same time, many of the messages that do get through are outright junk. This means you have to wade through page after page of dubious messages to get to the ones that tell you what you might need to do.

A replacement for the current e-mail standards could change all this. It would need to incorporate filtering, and to do so more effectively, it could include more data about the things to do found in the message. Already e-mail contains a few data items, such as the reply e-mail address and when the message was originally sent. In current e-mail standards, however, these are treated as metadata, not really part of the message, but information about the letter that the e-mail message supposedly contains. It would make filtering easier if these were treated as full-fledged data items and if there were more of them.

If you think of e-mail as containing mostly notifications, and few actual letters, you would want to add data elements that describe the follow-up action, its deadline, a link to the web page where action might be taken, and a link to a web page that explains the message. Then you would want your e-mail program to let you change these after you receive the message, in case your interpretation of what you need to do is different.

With a few more data elements in a message, the “letter” part of many messages could be eliminated or reduced to just one line. This would have multiple advantages. You could filter these messages accurately — when a message’s links go to a known web site, then your filters can prioritize the message appropriately. And with deadlines or due dates available to the filters, you can have more assurance that you’re not missing deadlines that are buried in your e-mail inbox. The link data elements would not be able to disguise a link with text that says something like “click here,” a favorite trick of spammers. Indeed, if an e-mail message contained a letter with a link in it, you would know it was probably junk. And without the ability to link deceptively, it would be almost impossible for spammers to send effective spam messages.

Much of this could be done within the current e-mail standard, but I believe it would be better if a new e-mail standard were drawn up. This new standard would be based on these ideas:

  • Treating messages as notifications, rather than letters
  • Filtering at the center of the standard, not an add-on
  • Adding data elements suitable for notifications
  • Treating data elements as data, rather than metadata, and a letter, if one is included, as just another data element
  • Treating links as data elements, not requiring them to be included in the letter
  • Incorporating identity information from participating social networks

With these changes, I believe e-mail could be made twice as productive. That is, you could finish the key things you depend on e-mail for in half the time. Of course, much of the time people spend in e-mail is spent on memos and documents sent through e-mail, and this won’t change that process, but it seems to me it could speed up virtually everything else that has to do with e-mail.

While it would be useful to have a standard that the world could agree on, there is nothing that actually forces anyone to wait for that standard to be put together. Any two social networks that wanted to get together on this could get the new e-mail system going, and since e-mail is pretty simple compared to the other things that social networks do, they could probably have something ready for users to try out within a month or two.