MIME-Version: 1.0 Server: CERN/3.0 Date: Sunday, 24-Nov-96 22:02:56 GMT Content-Type: text/html Content-Length: 13199 Last-Modified: Tuesday, 27-Feb-96 21:52:30 GMT Creating an Efficient Market on the World-Wide Web

CREATING AN EFFICIENT MARKET ON THE WORLD-WIDE WEB

Ramin Zabih

Cornell University

Consumers often purchase mass-produced items that are available from many different vendors. These different vendors usually charge different prices. If it were easy to compare prices, consumers would benefit substantially, and every consumer would be able to get the best price. This would create what economists call an efficient market. Until recently it was too expensive to make pricing information widely available. With the World-Wide Web, the cost of publishing information has fallen dramatically. Because of the Web, it may now be possible to create an efficient market for many mass-produced consumer items.

As an example, consider the market in computer peripherals (such as printers, modems, etc). In December 1995, the cheapest nationally advertised price I could find for the Hewlett-Packard HP660C printer was $330, while the most expensive price was $490. But it took many hours of reading advertisements to obtain this information. There are several hundred vendors who sell mail-order peripherals, and it isn't easy to compare prices.

Several efficient markets exist, including the stock exchange and the commodities market. Items on these markets (such as Hewlett-Packard stock) are available for a single price at any given time, and that price is widely known. Consumers rarely benefit from these markets because they do not trade consumer goods.

The PriceWeb Experiment

Together with some friends at Cornell I have begun an experiment called PriceWeb (http://www.priceweb.com). PriceWeb is an attempt to create an efficient market in computer peripherals. For a given peripheral, PriceWeb's web site provides a list of nationally advertised mail-order prices, listed in increasing order. This makes it simple to find the lowest-priced vendor.

If PriceWeb becomes widely used, it will have a major impact on the way that computer peripherals are purchased. Mail-order vendors would have to charge a single, uniform low price, unless they can offer customers some additional value. Computer peripherals are sold with fairly standardized warranties and return policies; vendors compete primarily on price and availability. This makes product differentiation difficult.

In an efficient market, most vendors would have to lower their prices. Short term, PriceWeb works to the advantage of consumers and to the disadvantage of most vendors. But in the long term there are advantages for vendors as well. Consumers will be able to buy items without worrying that they are over-paying (this argument is similar to Saturn's no-haggling policy for new car buyers). And, of course, prices should fall. These effects should lead to more customers

An on-line service like PriceWeb could have additional advantages for vendors. For instance, sometimes vendors wish to clear out inventory by selling an item very cheaply. (See Onsale for an example.) On-line pricing information can eliminate the delays involved in print media advertising, thus allowing vendors to more rapidly adjust their prices. Also, the Internet users who monitor pricing information will presumably be quite price-sensitive, so listing an item at a low price should generate a large response. These advantages to vendors have to be balanced against PriceWeb's negative effects on vendor margins. If the negative effects are larger, there will be consolidation among vendors.

How much would consumers save in an efficient market? One might expect mail-order prices on peripherals to be nearly identical, but they aren't. The average price spread on items that PriceWeb covers is $130, and some items have large price spreads. For example, the HP OfficeJet is advertised for as little as $377 and as much as $769.

If the vendors wished to avoid an efficient market, they would have several options. For instance, each vendor could sell slightly different items. If every HP660C printer were significantly different from every other HP660C printer, it would be impossible to compare their prices. But this would increase manufacturing costs and confuse potential customers, thus reducing HP's market share. This effect is illustrated by the American automobile industry, which for years offered an enormous range of options, partly to prevent consumers from comparing prices. The Japanese offered very few option packages, which gave them a price advantage in manufacturing. In the computer industry, almost everything is mass-manufactured, due to extensive standardization (see TNO 3(1) for a discussion of the economic impacts of standardization). So this option is not viable, because of opposition from the manufacturers.

An efficient market could also be thwarted by punishing the lowest-price vendor. For instance, suppose that all the higher-priced vendors refuse to carry Hewlett-Packard printers unless the lowest priced vendor is forced to raise its prices. HP could in turn raise the price that it charges that vendor, or cut that vendor off completely. But this probably would not be in HP's interests. More importantly, such actions would violate a number of laws against price fixing, such as the Robinson-Patman Act (an excellent overview is available from the Federal Trade Commission). These laws make it illegal for a manufacturer to fix the price that vendors charge.

An on-line efficient market would also be easier for the Government to monitor. Price-fixing in efficient markets is more obvious than in markets where there are multiple prices. For example, suppose a vendor cut its prices below costs to drive rivals out of business, and then raised its prices. This pattern would be fairly easy to spot in an efficient market with pricing information available on-line. The recent SEC investigation of NASDAQ trading practices was motivated by an academic study of pricing trends in this market; if this data were not available electronically, anti-competitive behavior would be harder to detect.

Challenges

There are a number of challenges that will need to be overcome in order to create an efficient market. One potential difficulty concerns updating price information. PriceWeb's information comes from printed national advertisements. In a market where prices change rapidly, advertised prices may be out of date before they appear in print. The obvious solution would be to use on-line prices from vendor's Web sites, but there are some obstacles. While a few vendors (such as Computability) provide on-line pricing information, most do not -- in fact, surprisingly few vendors have Web sites. Also, a price that appears in a printed ad comes with legal protections against false advertising. These guarantees may not apply to Web-based advertising.

Even if most vendors put their pricing information on-line, it may be hard to compare prices. For instance, consider Andersen Consulting's BargainFinder, which helps consumers buy audio Compact Discs cheaply. BargainFinder queries 9 on-line vendors that sell CD's via mail-order. Unfortunately, 3 of these vendors currently block BargainFinder from accessing their sites. Because BargainFinder actively queries on-line Web sites, it exists at the pleasure of the vendors. In the short term, they may not wish to cooperate.

PriceWeb will need to become economically self-supporting, which is a challenge. On-line pricing information is what economists call a public good (like, for example, a lighthouse). It is notoriously difficult to get the beneficiaries of a public good to pay for it. In addition, the Internet's culture makes it hard to charge users for information. I believe that consumers will not use PriceWeb if they have to pay for it. There are companies that think consumers will pay for such a service -- CyberSave Shopper, for example. If these companies are correct, they will create a market in pricing information. Mark Casson's chapter in Information Acumen (Routledge, 1994) discusses the effects on markets of information. An efficient market requires perfect information; as the costs associated with information fall, the market can become more efficient. There are costs associated with obtaining pricing information for computer peripherals. PriceWeb does not reduce the costs of obtaining the information, but instead collects the information centrally and makes it freely available.

PriceWeb might be viable simply because the cost of running the experiment is so low. Placing information on a Web page costs almost nothing, and the costs associated with gathering and entering data are also small. Even a small amount of revenue (from advertising, for instance) could make PriceWeb self-supporting. In addition, there are a number of services that PriceWeb could provide for vendors that would provide revenue. For instance, vendors could be automatically notified if a competitor beats their advertised price.

Will the Web Eliminate Retailers?

A web-based efficient market could also affect the relationship between vendors and manufacturers. Most manufacturers do not sell directly to customers, but via retailers. Retail stores carry products from multiple manufacturers, so a customer can compare products side by side. Manufacturers attempted to protect their retailers by various means; for example, the Manufacturer's Suggested Retail Price is usually quite high, so that every retailer can charge less than MSRP. Manufacturers protect retailers to ensure that retailers carry their goods. But with the growth of mail-order, the value provided by retailers could diminish. It may become profitable for the manufacturers to enter the mail-order business themselves. An efficient market could eliminate many middlemen.

An efficient market in consumer goods would have a substantial impact on society. For example, it could eliminate many retail stores in favor of mail-order vendors, who can locate in areas with low fixed costs. This could have massive consequences for many communities, which have downtowns built around retail shopping. Of course, consumers shop for other reasons than obtaining the lowest price. For some items (such as Tylenol), the savings may be so small that it isn't worthwhile to find the best price. However, large supermarkets spend a great deal of effort trumpeting small price advantages. It is possible that many shoppers are price-sensitive, and would prefer an efficient market. If the Web creates such a market, its impact could be far greater than any effects we have witnessed thus far.

While many retailers might disappear, some new businesses will also come into existence. For instance, it will become much cheaper to establish the kind of specialty businesses that traditionally only flourish in large cities. A business that only appeals to a small percentage of the population can flourish on the Web, because it can be easily reached by potential customers.

Conclusions

There have been several experiments in Web-based consumer empowerment, which are similar in spirit to PriceWeb. One example is the BBN Auto Mechanics List, profiled by Rich Lethin in TNO 2(8). This site lists local auto mechanics in the Boston Area. Besides providing names and addresses, it also includes customers' comments on their experiences with each shop. The comments are summarized by the moderator to provide an overall grade for the repair shop. A potential customer can view the comments of other customers, and can also add their own comments. Another example is The List, which performs a similar service for Internet Service Providers. Like Consumer Reports, these sites focus on giving consumers useful subjective information.

The Web has made an efficient market in consumer goods possible, because it has dramatically lowered the cost of publishing information. Until recently, only large corporations or a few unusual individuals could reach a large audience. The Web has changed this, and we are just beginning to understand its implications.