12 Garbage In, Garbage Out

1 2 3 4 5 6 7

James Hacker (prime minister): The statistics are unarguable. Humphrey Appleby (cabinet secretary): Statistics! You can prove anything with statistics. Hacker: Even the truth.


--From Yes, Prime Minister, by Jonathan Lynn and Anthony Jay


In August 1995, when the first paid-for advertising banner appeared on the original Web indexing service, Yahoo!, someone posted a memorial notice to Usenet: Yahoo! had gone commercial, and the Net was Dead. At the time, one of the most significant concerns on Netizens' minds was that sweeping commercialization of the Net was going to wipe away the existing culture by taking what was a free, public resource, sticking a meter on it, and selling it back to the people who had created it.


The Web's potential as a commercial medium was evident already, even though no systems were yet in place for handling secure transactions or heavy-duty product databases. It all happened very fast. In late 1995, you could browse a couple of pictures of flowers on a Web site and place an order by telephone; at the end of 1996, you could buy from Land's End's entire catalogue with intelligent ordering that told you if an item was out of stock and transmitted your order using encryption built into both your browser and the company's server. People want to do these things: half of everyone I know is desperately envious because in late 1996 a major supermarket chain started trying out selling groceries over the Net for next-day delivery in my particular residential area (the other half can't imagine why anyone would want to buy food this way).


Simultaneously, the Web's dominant nature has morphed: in 1994, most information sites were built by academic institutions or amateurs and hobbyists. By the middle of 1996, almost every media company was scrapping to set up a professional-quality news service, and journalists, who in 1994 were told they were about to become extinct, were being paid to supply these sites' content. There are no records of who was first to get paid to write for the Web; I can only say that when I started writing for d.Comm, the Economist's Web-based newsletter, at the end of 1995, I figured I was probably in the first fifty.


This complete reinvention of the Web brought with it a new set of clashes. Consumers in general don't want to pay for content on the Net; they want to be able to browse freely. Things may be different when Web-based services have been around longer and people know what they'd be buying, but in general there are too many unknown quantities for people to be willing to fork out for a subscription. There are exceptions; the Electronic Telegraph reports that people will pay small amounts for access to specialized pockets of content such as fantasy football leagues. The dominant answer so far is advertising sponsorship on a business model similar to that of radio and TV (as opposed to print, where being able to charge for the publication offsets some of the costs).


But advertising sponsors want the one thing most Net users don't want to give up: detailed demographic information. Net users hate this, partly because we're all so conscious of the amount of junk mail and junk telephone calls we get; the last thing we want is to be pestered over the Internet, too. Over time, actual sales sites have an advantage here, since you can't really order a new anorak from L. L. Bean without giving them at least some information about yourself, such as color preference, size, and shipping address. But if you're a business considering making the investment--anywhere from tens of thousands to millions of dollars--to set up a Web site, you want to know the make-up of the population that's going to use it. Cue another major cultural shift.


Most humans like to count things. Since the people using the Net in the early days were computer people, what they counted was primarily machines. In the excitement of watching the Internet spread like mold on damp bread, they counted the number of Internet hosts (machines or networks attached to the Net) and the countries in which they were located. The famous 1988 Internet Worm, which temporarily paralyzed part of the Internet and inspired a wave of computer crime legislation and hacker crackdowns, was a botched attempt to map this spread.


But companies don't market to machines (however much we wish they would); they want to count people. At least, that was the thrust of an argument constructed by


    

Copyright © 1997-99 NYU Press. All rights reserved.
Reproduction in whole or in part in any form or medium without written permission of New York University Press is prohibited.


Be sure to visit the NYU Press Bookstore

[Design by NiceMedia]