What happens when the world's best real-life bookstore takes on the world's best virtual bookstore?

It's located in the space once occupied by a hard-pressed auto dealership. The unadorned front window sports a haphazard display of dog-eared copies of such masterpieces as The Pantyhose Craft Book and Mystery of the Ductless Glands. Inside, the wooden floors creak as you walk past bookcases of new books shelved side by side with used ones.

Hard to believe you're in what is considered by many to be the world's best bookstore. Yet Powell's City of Books is a landmark for hard-core book lovers not only in Portland, Oreg., but throughout the United States and abroad. Even the likes of Barnes & Noble and Borders haven't been able to make a dent in its fanatically loyal trade.

Nevertheless, Powell's faces a competitive threat that many believe casts a cloud over the prospects for any bookstore's long-term existence. The business spearheading that threat is located up the coast, in Seattle--though it also can be said to be located everywhere and nowhere. The culprit is, the Web-based cyberbookstore whose explosively growing sales have called into question the most fundamental notions of how books--and perhaps other goods--should be sold.

How can a growing company that has mastered the art of doing business in a 3-D world respond to an incursion from a Web-based competitor? The obvious answer is that it should build its own beachhead on the Web. But extending a successful real-world presence into cyberspace can call for a very different approach from starting a new business on the Web.

By any measure, the book trade is amazingly fragmented and decentralized. Perhaps 50,000 titles are published every year in this country. Covering every subject imaginable, they are produced by hundreds of publishers, ranging from multibillion-dollar outfits like Disney and Time Warner to tiny shops like Graphics Press, of Cheshire, Conn. Those hundreds of publishers use hundreds of distribution companies to get their products to market. And none of this considers the huge market in used books.

As a result, book retailing is ripe for consolidation. Today the 800 or so superstores owned by Barnes & Noble, Books-A-Million, Borders, and Crown control almost 50% of all bookstore sales. As a result, hundreds of independent bookstores have closed.

Powell's staved off the threat from Borders and Barnes & Noble by turning itself into a store like no other. It was founded by Walter Powell, a retired painting contractor and carpet layer with no experience in the book business. Walter's son, Michael, went to graduate school in Chicago, where he opened a used-book store in 1970. When Michael went on vacation, he asked his father to look after the store. Walter enjoyed the experience so much that he went home to Portland and opened his own used-book store one year later. Not knowing anything about the book trade, he broke all the rules.

Instead of limiting his inventory to titles that would sell quickly--standard practice in most small bookshops--Walter bought books like crazy. He built up the stock until it swamped his original location in the northwest part of town and then a second space across the street. He finally pushed the whole operation a few blocks away, into a defunct car dealership that he converted into what is now the single biggest independent bookstore in the United States and perhaps the world. "The idea is choice," says Michael Powell, who joined his father's business in 1979. "We went way past the traditional size of bookstores to find the limit of what people would be interested in, and we've never found that limit."

Today Powell's City of Books is like the Platonic ideal of funky independent bookstores. It rambles over 43,000 square feet of retail space and carries more than half a million volumes in 122 main subject areas. It has a laid-back, paper-strewn cafÉ that sells espresso, pastries, and about 1,000 periodicals. In the 1980s Michael Powell established six satellite stores, all in Portland, most of which specialize in subjects like travel, cooking, and computers. Altogether Powell's has some 400 employees; sales top $30 million a year.

Much of the company's continuing growth stems from a second innovation: selling books in every form--hardcover and paperback, new and used--on the same shelves. "In the used-book world, mixing paperbacks and hardbacks was not so much of a stretch," Michael Powell says. "But when my dad had the idea of bringing in new books, too, I had no sympathy. Used and new on the same shelf? It seemed crazy." It wasn't, though. Combining books in every form, Powell says, "had synergy way beyond what we expected. If you put all the new books in one store and all the used books in another, each wouldn't get half the total business--they drive each other."

Providing books in all forms, Powell's is able to offer a selection greater than any of the chains. "It's our weapon against Borders," says Kanth Gopalpur, marketing manager for Powell's. "They're an excellent store. They create a nice ambiance and have a good cafÉ, and so do we. But they don't sell 21 editions of The Catcher in the Rye. We have old paperbacks that sell for $2, a signed first edition for $1,200, and everything in between."

Michael Powell wanted to keep growing the business, but he didn't, as he puts it, "have the personality" to enjoy rolling out stores. Even if he had wanted to, the logistics of creating another City of Books in another city were formidable. Expansion didn't seem to be in the cards. Then came the Web.

Meanwhile Jeff Bezos had been thinking about the Internet and books. Bezos had always enjoyed bookstores: the smell of paper, the quiet sense of promise behind every spine. But he had never considered starting a bookstore until the early 1990s, when he, like many others, began pondering the commercial implications of the Internet. Bezos was a senior vice-president for a Wall Street fund, but he wasn't so far removed from technology to realize that to sell anything on a medium as emotionally cold as the Internet, he would have to offer a product that was so superior that customers would brave the obstacles of the technology to get it.

The key, Bezos decided, was to sell products in so many formats that no single place could offer them all. He drew up a list of 20 possibilities, including magazines and PC hardware and software. But he settled on his early love: books. Even the biggest physical bookstore can't purvey more than a fraction of the 50,000 books printed every year. But a virtual bookstore could, in theory, offer every book in print. Bezos decided to call his bookstore Amazon--a mighty river of text. With the Web address, it opened in July 1994, offering more than a million books. (It now offers more than 2.5 million.)

Physically, Amazon is about as different from Powell's as it can be, while still being a gigantic independent bookstore. Whereas the City of Books in Portland has the down-home atmosphere of an old country barn, Amazon, in downtown Seattle, is based in a modern office building. Actual for-sale books are nowhere in view. The Amazon office is characterized instead by the paraphernalia of rapid business expansion: cardboard boxes stacked in hallways, cables draped across the carpet, desks made out of doors and two-by-fours.

In marked contrast to Powell's, Amazon stocks only several hundred top-selling titles, all of them in a warehouse in south Seattle. The store functions almost entirely as an intermediary, taking customers' requests on its Web page and pumping them to distributors, who fill the orders. The books are shipped back to Amazon; from there they are sent out to customers.

Bezos believed that Amazon would not succeed unless people enjoyed visiting its Web site. "We had to make it a destination," he says, "the way any enjoyable store is a destination." Which meant, in a sense, providing as close to a Powell's-like experience as he could in a medium that is inherently slow and tedious. Amazon's Web site cannot provide a cafÉ, friendly clerks, or satisfyingly overstuffed shelves. It doesn't even offer fancy graphics and animation. But it does provide a simple (by Internet standards, anyway) tool for searching its catalog, amusing contests, chatty lists of best-sellers and award winners, interviews with authors--and up to a 40% discount on books reviewed in the New York Times Book Review, the New Yorker, and other publications.

Amazon also provides services that regular stores don't. People who search for a book at the Amazon site often find a description accompanied by excerpts from reviews--not only from print sources but from customers. Authors, too, are invited to comment on their own works. And Amazon asks customers which kinds of books they like. When books in the same category or by the same author appear, the company sends E-mail inviting people to buy them.

To put together the business, Bezos faced the difficult problems of assembling reliable, user-friendly software and negotiating with the book industry's hundreds of distributors and publishers. But perhaps his greatest worry lay in another direction. Like Michael Powell and independent bookstore owners everywhere, he was peering over his shoulder at Borders, Barnes & Noble, and the other giants. The chains, too, had the muscle to deal with all the distributors. They, too, could offer every book in print. Whereas Powell's could distinguish itself by offering a unique range of wares, Amazon could not. Thus Bezos had to rely on the ancient strategy of getting there first, with the most to offer.

"Most people remember two or three brand names for any given product," Bezos explains. "In shoes, there's Nike, there's Reebok, maybe Adidas, and then there's everyone else." In the real world, he argues, the problem can be avoided to some extent, because when customers go to a store with an unfamiliar name, they can see merchandise in the windows and lines of cash registers inside. But because no Web page can provide that tangible reassurance, brand names are especially important in on-line commerce. Customers are likely to visit the two or three sites they know, rather than try to find and evaluate new ones. As a result, Bezos says, "I thought, Well, Barnes & Noble and Borders will set up Web sites, and everyone already knows their names. So that leaves maybe one open slot for everybody else on earth."

To secure that slot, Bezos had to work fast: he was by no means the only budding literary cyberspace entrepreneur. To break free of the pack, he turned to venture capital, a business tool more common in Silicon Valley than on publisher's row, and he didn't get rolling until he had a kitty of "a few million dollars." When order money began to come in, he plunged it back into merchandising, advertising heavily both on Web sites like Slate, the electronic magazine from Microsoft, and in print publications like the New York Times Book Review.

The strategy worked. Sales in the first couple of years of Amazon's existence rocketed from zero to $15.7 million last year. And in the first quarter of 1997 alone, sales reached $16 million.

Amazon's success has been so complete that the inevitable questions have been raised by virtually everyone who thinks about the book business: Does the convenience of ordering books on-line doom the physical bookstore? Or do the old and new ways of doing business complement each other?

Michael Powell wasn't seeking to level the playing field with superstore chains when Powell's began its on-line service. Nor did he look for venture capital--he spent, compared with Bezos, next to nothing. For Powell, setting up shop on the Internet was something of an offshoot of a different endeavor. The company's Web site was created in one of the more torturous processes Powell's has been through: computerizing its stock. In addition to the million-plus books in the inventory, the store buys 2,000 to 4,000 used books daily. "We had to get specially written software to control the constantly changing database," says marketing manager Gopalpur. "And then it took us two years of round-the-clock drudgery for data entry."

As soon as the scut work was done, he says, it was natural to take the digital information on-line. In 1993 the Internet site debuted, offering books from the company's computer-book store, which was the first to computerize its catalog. To Gopalpur's surprise, within hours on-line orders started coming in.

A year later, Powell's moved the catalog of technical books onto the Web, and by 1996 the complete catalog was on-line ( Since then, sales have grown by 10% to 20% a month. A back room in the office overflows with books going to New York and Iowa, Canada and Korea. In five years, Michael Powell says, the store may bring in a third of its sales on-line.

Powell says that he's "not putting any money into marketing." Rather, he's banking on the better mousetrap theory: build one, and the world will beat a path to your door. In this case, the better mousetrap is one of the world's biggest and best-organized stockpiles of used books. Like Amazon, Powell's has an almost graphics-free user interface, suitable for people with slow Internet connections. But Powell's doesn't provide reviews, contests, customer comments, or E-mail notifications about book arrivals. Instead it offers a simple way to search through a list of new and used books--especially used books, which, as Powell likes to point out, are books that people either can't buy anywhere else or can buy elsewhere at a higher price. Unlike customers in the physical store, on-line customers treat the new books as lagniappe--extras to toss in while ordering used books. The focus on used books allows Powell to welcome the presence of Amazon and other Internet vendors of new books. "They get people used to buying and ordering books on-line," he says. "That can only help us in the long run."

Aiming for the worldwide bibliophile market, Powell's is setting up its site in other countries. Its first foreign partner is a Japanese Internet-marketing company that has set up links to the on-line catalog. "If we find good vendors in each country," Powell says, "they can add our product to their mix at little cost except for the telecommunications expense." Powell says that people all over the world will eventually take the virtual voyage to Powell's "as long as we have something they can get nowhere else. You can make a bookstore a destination in the real world by having a wonderful atmosphere, a good cafÉ, and great service. But the Internet gives you fewer options. You have to be unique because you're competing with everyone else on earth."

Despite running bookselling businesses that are now a mere mouse click apart, Powell and Bezos aver that they're not competing. In fact, each insists that the other's presence on the Net will, if anything, help its business.

This equanimity may be more than the businessperson's usual caution about publicly bad-mouthing the competition. Both entrepreneurs have created something entirely original, and there may be room for both to thrive. On the other hand, real-world bookstores that don't provide a unique inventory and a stimulating experience--as well as virtual bookstores that don't offer an extraordinary catalog and tremendous convenience--might well wonder about their future.

Web Design
Elaborate Isn't Always Better

When Michael Powell decided to take Powell's City of Books on-line (, he chose not to imitate Amazon's successful and elaborate virtual bookstore ( Instead he set out to create a simpler site that played to his customers' no-nonsense desire to access a huge inventory of used as well as new books.

Here's how the two sites compare, along with Michael Powell's reasoning:

Focus New books Used books
Rationale: The company's niche
Scope Lists all books Lists only books on hand
Rationale: Avoids unfillable orders
Promotion Heavy Web promotion Little promotion
Rationale: Word-of-mouth more effective
Associated content Reviews, chats, articles, more Nothing but books
Rationale: Customers know what they want

Charles C. Mann is a contributing editor at Science and the Atlantic Monthly.

Published on: Jun 15, 1997