Forced to spend three hours of each twelve hour workday handling about 150 incoming messages, Steding thinks, "The volume of email is so high, it's become devalued. Sending an email is no longer sufficient to ensure communication. You've got to go see them nose-to-nose. It's like the old way, except back then you didn't have to send email first. We're going backwards."
The Information Revolution was supposed to save time and eliminate drudgery. What's gone wrong? Are we doomed to suffer from Info-overload from here to eternity? Or do we just need better tools for sifting info-wheat from info-chaff?
If such tools can be invented, they'll have to be aligned with the fundamental properties of information itself. Claude Shannon, the father of "communications theory," and the fellow who coined the term "bit" (binary unit), once said "Information is news that makes a difference. If it doesn't make a difference, it isn't information." A radio traffic report about a car crash up ahead is information if you can still change your route. But if you're already stuck in the backup, the message is a useless bunch of bits.
By Shannon's definition, most of the messages coursing through corporate networks aren't worthy of the name "information." Junk messages clog emailboxes, obscure useful information, and force workers to waste hours each day simply trying to keep up with the flow of messages.
Some observers blame the "group list" feature found on today's messaging systems. "Group lists are an incredibly efficient tool for disseminating information when it's relevant to everyone in the group," says Don Wood, director of marketing at Octel Communications, a leading voicemail systems supplier. "But the abuse of group lists causes a huge waste of time. Virtually all the junk I get comes from being on so many group lists."
In a way, the "group list" phenomenon expresses the best and worst aspects of the Information Revolution. In two decades, innovation has slashed the cost of moving bits about a million-fold. Today, the incremental cost of sending out one more copy of an electronic message is virtually zero.
Taking advantage of this remarkable technical achievement, everyone with email or voicemail access can be a micro-broadcaster, able to inundate colleagues at no cost to herself or her department. Almost overnight, the elimination of sending costs has reversed the fundamental economics of communication. Instead of costly transmission limiting the message supply, today's scarce resource is the attention span of recipients.
To counteract the "group list" problem and reduce time wasted sorting messages, software firms like Massachusetts-based Beyond, Inc. have introduced so-called "Bozo filters" that cull out unwanted messages. These programs prioritize incoming messages once users have entered the names of those whom they want -- and don't want -- to get email from. Octel recently introduced a similar feature for its voicemail systems. Presumably, a message from your boss will leap to the top of the stack, while the memo announcing a change to the company dental plan slips to the bottom.
As a first step, using a "Bozo filter" may be better than leaving your emailbox undefended. But in the final analysis, filtering software can't get to the heart of the problem. Since filters require you to preload the names of people from whom you want mail, they assume you already know what you want to know, or at least who you want to hear from. But "news that makes a difference" -- that flash of insight or speck of data that alters the course of a business or a career -- often comes from unexpected sources.
What's really needed is a quick, easy way of gauging the value of an incoming message. Since the sender is the only one in a position to guess whether a message might be of value to the receiver, the sender should label each message with a super compact indicator of value -- more commonly known as a price.
Prices are, after all, a wonderfully concise form of information about information. Everything known about a company's financial performance and prospects is reduced to one number -- its stock price. Instead of analyzing stacks of data comparing various products, we often rely on their relative prices to guide our choices. Price signals allow us to instantly assess relative values and focus our attention on the subtle issues that deserve careful scrutiny.
Imagine an email or voicemail system where the sender indicated her sense of the content's value by attaching "urgent, standard, or junk" levels of "electronic postage" on each copy of an outgoing message. The "postage" would be charged against her budget. Conversely, her account would be credited with the value of postage attached to messages she opens. With senders once again facing some marginal cost for each message they put out, they'd think twice before copying someone without reason. The profligate use of group lists would disappear, the volume of junk messages would plummet, and recipients would zero in on messages bearing the most "postage" -- and the most importance.
Though little known, a messaging system quite close to this approach was implemented a few years back at AMIX, a now defunct Silicon Valley startup. Alongside its normal email system, AMIX launched a market-based messaging system. Message headers like "$5 if you read this message," gave recipients an incentive to open their mail. Individuals accounts were credited with moneys debited from the sender's account.
Mark Miller, who helped design the system, reports that AMIX's conventional email system fell into disuse as people switched over to a market-oriented system that rewarded them for the time they spent handling messages. "Instead of treating email as a chore," says Miller, "people began seeing it as an opportunity."
Asked why the market approach to network management hasn't caught on yet, 3Com's Tom Steding replies: "The perception of the problem has to get to the point where people will want to do something dramatically different. Despite all the groaning, the cost of wasted time is widely distributed and unmeasured." True enough. But even a rough calculation shows that thousands of hours wasted sorting email adds up to big dollars.
Miller, now senior architect at Los Altos-based Agorics, Inc., has another explanation for the slow development of market-based messaging management. "Even with the emergence of markets in financial derivatives, the idea of markets in non-physical goods, in abstractions like 'a moment of your attention,' 'the right to pollute,' or 'the right to transmit on a given frequency' is still very counterintuitive. Besides, the naive assumption is that better technology will eliminate congestion."
But paradoxically, advancing technology only worsens network congestion. For example, with the coming shift from Ethernet to fiber, network capacity will jump at least 60 times, and make videomail feasible. The hitch is that "Two frames of video require more bits than a month's worth of email at most companies," says Miller. Though a short videomail may convey more information value to the viewer than the same message sent via email, each video bit carries radically less information value than each text bit. If the network treats all bits as equals, videomail will congest the fibernet and keep "old-fashioned" email from getting through.
Recognizing that the only way to intelligently prioritize messages is by allowing senders and receivers to bid for access to the network, Miller and his colleagues at Agorics are now developing a market-based fibernet management system that uses an internal software agent acting as a bandwidth auctioneer.
Just as today's networks unleashed a flood of low value per bit email and voicemail, fiber will soon make it possible to send ultra low value per bit videomail. Filters can't hold back this tide. No matter how much capacity is added to a net, its bit traffic will grow even faster. As system complexity evolves, message pricing will become the only sensible way to allocate the bandwidth that matters -- the bandwidth between your ears.