One fine day a young woman went out to buy a car. The dealer convinced her to purchase a Ford Taurus for US$19,500. She said she needed to sleep on it and would come back the next day. But instead of just sleeping on it, she used the Net to inquire whether there were others, near her, who were also considering buying a Taurus. By next morning, she had found 15 people who were. Some email discussion ensued, and she returned to the dealer to say she would take the car, but for $16,500. This was so far below his price that he assumed she had made a mistake. "No sir. I have not made a mistake," she replied. "I simply failed to mention that I am buying 16 cars, not one." Delighted with the idea of selling in such volume, the dealer promptly sold the cars at her price.

A buyers' cartel - as opposed to a sellers' - is almost impossible to create: too many people need to be involved. Meeting with, speaking to, calling, or finding those who may be interested is too difficult, and you probably wouldn't know who to contact anyway. Consumers of products find themselves in a poor position compared with suppliers, people who, by virtue of knowing one another, can fix prices anytime they agree. In addition, suppliers are typically few in number. The consumer's position is weakened because he or she cannot shop efficiently. The potential buyer cannot cover an area that is wide enough to be significant.

This is about to change.

Let your fingers do the walking?
In the days before Pattie Maes was a mom, and prior to joining MIT's tenure track, she had plenty of time to browse through stores, newspapers and magazines, even cities, with the hope of discovering some piece of treasure. These days she hasn't the time to explore at a leisurely pace, but even worse, the amount of information and the number of products have expanded almost exponentially. What was merely overload yesterday has become impossible for her today. I must say I have felt this way for years!

I don't even refer to the Yellow Pages. Instead, I trust friends, whose recommendations have proven to be on the mark. There is my sister-in-law for movies, Jerry Rubin for restaurants, and Sherwin Goldman for wine. Like most of us, I have few people to whom I can turn. Otherwise, I rely on critics, experts who provide evaluations. In other words, today we have only two choices: ask a friend or trust an expert.

Thanks to the work of Professor Maes and her colleagues, including former students who have started Firefly Network Inc. (formerly Agents Inc.), we now have a third way to find a new film, a hip restaurant, a timely news article, or a hot Web site. The concept is called collaborative filtering - a way to tap into other people's wisdom.

Have it your way
In July 1994, a program called RINGO became the first instantiation of this concept. It helped users find interesting music. Mind you, this is much more difficult than locating the lowest-priced Ford Taurus or the most inexpensive pair of Porsche sunglasses. Cost is easily measured, but "interesting" music is very much in the ears of the beholder.

What RINGO did was simple. It gave you 20-some music titles by name, then asked, one by one, whether you liked it, didn't like it, or knew it at all. That initialized the system with a small DNA of your likes and dislikes. Thereafter, when you asked for a recommendation, the program matched your DNA with that of all the others in the system. When RINGO found the best matches, it searched for music you had not heard, then recommended it. Your musical pleasure was almost certain. And, if none of the matches were successful, saying so would perfect your string of bits. Next time would be even better.

The idea is simple, but the existence of the Net makes it more than cunning. When graduate students first put RINGO online, thousands of users and music titles were on it within a week. The expansive and global nature of the Net takes word of mouth, as we know it in real time and space, and reaches across all time zones and the planet itself. Try it. The current, commercial incarnation of this idea can be found at www.ffly.com/.

Taking the noise out of the Net
Many users, old-timers and newbies alike, complain that the Net is too noisy. There is no easy way to separate the opinions of an 8-year-old from those of The Washington Post (not sure which I would prefer). Remember the old axiom "Garbage in, garbage out"? Well, it is just that - old. It's about time people realized that the noise of the Net is its beauty, and that navigating this noise will become easier and easier because of electronic word of mouth.

Until recently, the assumption has been that you must have brand recognition. Brand names, we are told, make information credible. If The New York Times says it, or if John Markoff says it, you'd better believe it. Otherwise, beware. Search engines don't discriminate between a high school term paper and Britannica's homepage, let alone the writings of Cliff Stoll and this back page.

Electronic word of mouth does. And it works both ways. It not only allows you to find music titles of obscure ensembles, for example, but it very quickly blackballs the bull. It means that one person's three-star restaurant can be an anathema to another. We have seen just the beginning of a new kind of Consumer Reports - done by consumers, for consumers.

For Pattie Maes and company, the ultimate effect of this technology will be demonstrated when, for instance, a band signs with a big label because Firefly generated so much excitement about its music. That is, when a new product will be launched because word-of-mouth technology formed an online cartel of people who want it to be sold.

Pattie Maes (pattie@media.mit.edu), a professor at the MIT Media Lab and founding chair of Firefly Network Inc., contributed to this column.

Next Issue: The Digital Absence of Localism

[Copyright 1996, WIRED Ventures Ltd. All Rights Reserved. Issue 4.10 October 1996.]