The middleman myth and "disinter-remediation"

The Internet was supposed to cut out the middleman, but you've said the opposite is happening. Can you explain that?
Conventional wisdom is a terrible thing. Conventional wisdom is generally wrong about everything involving the Internet. Four years ago, conventional wisdom was that the Internet was the death of advertising. Well, that was absurd! It was a full employment act for everybody in the advertising business.

The latest concept of conventional wisdom that is wrongheaded is called "disintermediation." The essence of disintermediation is somehow we're going to eliminate the middleman and go directly from buyer to seller, and everybody in the middle is going to be history. Real estate agents are gone, bankers are gone, etc. That is absolutely not what's happening. We are throwing out a lot of the middle players, but we're also putting in new middle players. The simple fact is that information technologies complexify the business environment. They create more options and opportunities and more niches for players. So what we're doing is creating this turbulent situation where we're throwing out some of the old players, we're putting in some of the new players, and at the end of the day we have more people in the middle, not less.

What's an example of that?
There's a historic example that demonstrates it. Once upon a time in the late 1950s, if you wanted an airline ticket, you went straight to the airline. You didn't go to a travel agent. You only went to travel agents for hotels and cruises and things. Along comes the mainframe computer and an executive at American Airlines has this really bright idea to build what would become the Sabre system. They needed someone to run this automated terminal system, so they gave it to the travel agents. Suddenly, you no longer went directly to the airlines; you went to the travel agent who was your intermediary to the airline. It was so effective and so good for American that the Justice Department said, "You have to let the other airlines on it." So Sabre, which was this unique advantage that American had, then became a commodity for all the airlines to use, but the travel agent was firmly intermediated in the mix.

So then another few years later, another executive at American Airlines said, "We need something that makes us unique. What can we do?" Minicomputers had arrived and they built the first frequent flyer program, AAdvantage, and once again mediated this new thing between the traveler and the airline for the airline's advantage. It didn't stop there. All the other airlines did the same thing and it became a commodity, still influencing purchase decisions. Then the credit card companies got in on the game using their computer systems, where suddenly you could buy a MasterCard or a Visa card that allowed you to get frequent flyer points.

Suddenly, what was a very simple relationship between traveler and airline was now a complex relationship, full of innumerable players. All of a sudden, people are doing things (thanks to computers) like arranging their flight schedules in ways calculated to get more miles, even though it was not the fastest route, and buying things on their credit cards for reasons that had nothing to do with travel, solely to get frequent flyer miles.

Information systems complexify. They create new niches, new opportunities for players. The more you put information technology in, the longer and more complicated the value chain becomes.

Yet I see the opposite. American Airlines sends me email every week telling me what their specials are if I buy directly from them.
Right. What's happened is that value chains have disappeared and in their place we now have value webs. The relationship between you and American Airlines, there are moments when it is direct (like if you're talking to the frequent flyer program), but there are other times where it's indirect and accelerated by your travel agent saying "Gee, American has a special. Do you want to use that?" They do promotions to the agent, the agent promotes to you, and it influences your buying decision.

To the question of how long can a value chain be, the answer is identical to something that Benoit Mandelbrot, the father of fractal geometry, wrote in 1978. There's an article he wrote titled, "How Long is the Coastline of England?" Well, the answer is a) it depends, and b) effectively infinite. You see, it depends because it's one length if you're on a plane flying around the island; it's longer yet if you're on a boat going in and out of little coves; it's longer still if you're in a car driving along the bluffs; and longest of all if you're an ant going up and down each one of the pebbles on the way around. It's effectively infinite. Well, how long is a value chain? A) it depends, and b) effectively infinite. Sometimes it's very direct, but other times it's very long and circuitous. And the more the computer is in the mix, the longer it becomes and the more circuitous.

People who think that there is disintermediation happening have got a very static view of things. We are throwing out middle players today, but we're going to put new middle players in. It's really a process of "disinter-remediation." The computer comes in, it changes the commercial environment, the middle players that do not adapt to that change are tossed away, and new players that can deal with the new environment come in and assume roles.

This isn't anything new. Information systems have been doing this for as long as anyone can tell. Exactly the same thing happened thanks to the printing press in the closing years of the 1400s. The printing press and both direct and indirect impacts led to the invention of modern mercantile capitalism between about 1490 and 1520. In fact, you can trace it to the publication of a little book called The Treviso Arithmetic in 1476. Published in a suburb of Venice, it's the earliest known dated book of shopkeeper's math, in vernacular, not Latin. And it probably wasn't the very first, but it's the earliest one that we know of. Think about what you need for capitalism. You need a class of numerate shopkeepers that can do the business math. This was a how-to book of business math. Out of this innocent little tome grew everything that we take for granted today.

What sort of implications does this myth of disintermediation have for businesses?
There are a couple of implications for businesses about disinter-remediation. One is do not blindly think about getting closer to your customer. Sometimes effective communication means getting farther away in the right sort of way.

If you're selling stereo systems, you want to make sure that the customer has your product brought to his or her attention at Circuit City. You need to find the right kinds of partners. So the nature of this business is picking partners right. Then second is accepting the fact that partner relationships are going to be as volatile as the marketplace.

Once upon a time, it was enough to pick the right partner. Now, you have to pick the right partner in time and effect the partnership quickly enough so you can take advantage of the market, and also be prepared to break the partnership on friendly terms when it no longer makes sense. You don't want to be mean to partners, but everybody walks in saying, "OK, here's this moment in time. We have an opportunity; we'll go after it. The moment the opportunity disappears, we go our separate ways, and we may come back together again when conditions change."

The other element of this--I think this is very good news for entrepreneurs--information technologies scale badly. When an information technology first arrives, it's easier for a small organization to use it than a large organization. Because information technologies scale badly, it means the advantage adheres to the small players. It's an unlevel playing field and the small player has the advantage over the big corporation. Big corporations have money and resources, but they can't use new technologies as effectively as small bands of entrepreneurs with crazy ideas, shoestring budgets, and not a lot of adult supervision.

Is there a Web example of that?
There's a classic Web example, and that is Amazon.com vs. the established bookstores. [Amazon.com founder] Jeff Bezos had a fabulous inspiration. He was not a book man; he was a computer guy on Wall Street and said, "This is a fabulous vehicle for selling things. What would people like to buy and sell?" He looked at the demographics and said, "This is a no-brainer: books." He looked at the book distribution and said, "Really terrible distribution system. Crown Books is appalling. All you get is the main stuff. Can't get all those other books. I've got publishers who want to sell small run books to the public. Let's connect and create a more efficient marketplace...and then do creative things."

The idea on Amazon.com is that if Nicholas Negroponte has a new book out, Amazon.com sells it, and on my Web page I can say, "Here are my five favorite books right now. One is Nicholas Negroponte's Being Digital," and at the bottom there's a button that says "buy it." You click that button, you buy Negroponte's book through Amazon.com, and I get six percent of what you pay. That's a virtuous cycle. So it's a whole different way of buying and selling stuff.

Amazon.com got in there ahead of everyone else because it had the vision and also could scale with the technology. Now the big boys are coming: Barnes and Noble comes in, has a nice Web page, but did the classic big-company stuff. Really dirty trick with the lawsuit it did right around the public offering. To me, that crossed the line. That was a bad business practice...and it's smarter than that. That was just mean. But it got back to business. Now it's doing a Web page and chasing Amazon.com. Will it catch up? Who knows? There's space for more than one bookstore in cyberspace.

NEXT: Free market determinism and other follies