People say the customer is always right. "Don't believe 'em," warns Jim Campbell.

This article originally appeared in Inc. Technology magazine

Campbell is president of P.W. Campbell Inc., a $35-million Pittsburgh company that designs and constructs buildings for financial institutions such as banks and credit unions. He listens to his clients' ideas--for a fountain in the middle of the lobby, say--and then determines whether or not they're good ideas. "Even if we can do it, if we know it's not right or economically feasible, we'll tell them," says Campbell matter-of-factly. "Otherwise, everyone has wasted their time."

Campbell's search for a Web-development firm reflects that philosophy--only in reverse. Knowing next to nothing about Web-site creation, the president went looking for someone who wouldn't do his bidding as much as explain to him what that bidding ought to be.

Campbell first considered launching his 88-year-old family business onto the Web back in 1996. An America Online member, he had spent enough time fishing around the Internet to conclude that a presence there would send the right message to his technologically savvy clients. He also liked the idea of being out front: while plenty of national construction companies had sites, none of his regional competitors did.

Campbell had a general idea of what he wanted: the prestige of a URL on his company's letterhead, a place to post his quarterly newsletter, a way of offering information on building trends in the financial-services industry, and a total price tag not exceeding $15,000. Beyond that he needed guidance. So he spent the early part of 1997 hitting up colleagues, business associates, friends, and relatives. "Whether people had a site, were considering getting one, or knew nothing about it, it was my target topic of conversation," he recalls.

Most of those consulted were as inexperienced as Campbell was. But he did pick up the names of 10 development companies, which he passed on to an employee for an initial round of screening. He then established three hard-and-fast criteria for choosing among the Web designers. First, serious candidates had to be able to handle all aspects of the project, from designing and programming the site to registering a domain name to hosting. The last thing Campbell wanted was to bounce around among vendors. "Even if we paid for it, it was going to be worth it," he says.

Second, contenders had to be focused on the customer, not on themselves. Providers that launched into a spiel about what they could do as opposed to what they should do specifically for P.W. Campbell were out. "Some didn't even ask what our business was," says Campbell. Others kept pitching their advertising services even after Campbell had assured them that all his marketing needs were covered.

And third, prospective vendors had to be willing to travel to Pittsburgh for face-to-face meetings. Although he knew Web sites could be created virtually, Campbell wanted to work closely with the provider. In addition, he was not in the habit of doing business with someone whose hand he couldn't shake.

Those criteria winnowed the list to three contenders. Then Campbell popped the trick question. The president asked all three companies to provide a proposal. Two declined, requesting a sit-down meeting instead. They stayed in the running. The third sent a proposal. It flunked. Campbell suspects that a meeting with that vendor would not have been productive anyway. "All they asked were questions like 'How many pages do you want? How many links do you want?'" he says. "They were looking to take a cookie-cutter order: Tell me what you want, I'll add it all up, and here's what you get."

Sitting down with the first of the two survivors, Campbell launched into a lengthy explanation of his business and his goals for an on-line presence. The feedback was negligible. "Everything we asked, no matter how big or how small, we were 'yes'-ed to death," he recalls. In addition, the vendor's marketing materials didn't inspire confidence. "We'd never let something that messy and unprofessional go out of our office," Campbell says. "And they were supposed to be a design firm?"

But things were different when Campbell met with Bruce Colgan, the CEO of Interteq, a three-employee firm based in Pittsburgh. "We asked him all the same questions," says Campbell. "But instead of just saying yes to what we wanted, he would say, 'Yes, you can do that, but I don't recommend it, and here's why.'" For example, Campbell was hot to use 3-D panoramic images on the site, but Colgan warned him off them because their creation would delay the launch. Colgan was also comfortable talking money up front--a subject the other company had avoided. "Somewhere along the line they needed to say, 'Define expensive for us,'" Campbell remembers thinking when he met with the other firm.

Although Interteq had the edge, Campbell still wanted to compare the companies' proposals. The first included a needs assessment priced at $500 to $1,000, an amount that would have been applied to the project fee if the firm had been hired. Campbell says he wouldn't have minded paying for the assessment, but he wasn't thrilled with the company's references. They were all good, but they were also from technology departments, something P.W. Campbell doesn't have. "They all talked the same technical language," Campbell says. "I didn't. And I didn't want to."

Interteq's proposal, however, was beautifully detailed and included a step-by-step outline of the project: define the site's goals and audience, create an easy-to-navigate structure, design the pages, place the site on a server, establish a domain name, and register the name with search engines. Colgan estimated the cost at $3,000 for 50 hours of labor--well below what Campbell had expected to pay. Once its references checked out, Interteq was hired.

For the next six weeks Colgan and his two Interteq staffers methodically worked their way through the outline. Campbell chose to stick with his existing marketing materials but edited them extensively after Colgan stressed the perils of text-heavy pages. "He told us, 'You're not trying to sell anyone here. You're just showing credibility,'" Campbell says. "We had to keep remembering that."

Throughout the project, Colgan alerted Campbell to new developments on the site via E-mail, generally embedding URLs in his messages so Campbell could click directly to the pages in question. Campbell would either OK the work or request a little massaging--for example, the addition of a one-page table of contents to better acclimate readers to his company's newsletter. Colgan was also a fount of advice. "Bruce didn't come to us and say, 'Here are three ways this page can look. Which do you prefer?' He said, 'This is what I think looks good.'"

Last, Interteq made arrangements for Campbell's site to live on the server of a local Internet service provider (ISP) for $10 a month. The site was launched last May; since then Campbell has received several solid business leads from people who found him on-line. He has also learned that potential customers like to check him out on the Web before they call--essentially taking some of the fear out of a blind date. Anxious to impress prospective as well as existing customers, Campbell has now begun jazzing up his pages with technologies he eschewed back when time was an issue. First up: those 3-D panoramic images.

No More Funny Business

The first time Rob Snell tried to put his company on the Internet, he failed.

Snell had founded Gun Dog Comics while he was a college student, in 1988. By 1995 the half-million-dollar business had four locations in and around Starkville, Miss., but sales had begun to slump. Snell knew he had to find a market beyond that of the comic-book aficionados of Mississippi State University. He thought he might find that market on the Web.

The fulfillment part of electronic commerce didn't faze Snell in the least. After years of helping out at his parents' pet-supply company, he understood mail order. To further aid matters, he was already a credit-card merchant, and he didn't have some elaborate database that he'd have to tie into an electronic front end. Snell felt that all he needed was a simple site where customers could select items from his inventory and drop them into a shopping cart. The shopping cart, in turn, would generate an E-mail message telling him how much to charge the customer and where to ship the merchandise. "I just needed someone to handle the Internet side of the street," he says. "I was prepared to take care of everything else."

Still, before calling vendors Snell decided to embark on a little self-education tour. Typing such phrases as "E-commerce," "mail order," and "small business on the Internet" into various search engines, he began by reading articles and general advice about fielding commercial sites. He then visited successful retailers like Amazon.com and checked out other mail-order businesses and hobbyist/collectibles sites. On several occasions he even bought something on-line in order to run a site's order-taking process through its paces.

The results were not impressive. "I was blown away," Snell says. "Fifty percent of the orders I placed were either incomplete or I never got them. There was no E-mail confirmation. No 'Thank you for your order.' Nothing." Moreover, Snell often found placing an order to be too cumbersome, with too many clicks required to find the shopping cart.

Nor was Snell bowled over by the Web-design firms he found as a result of his research. Most didn't offer commerce services at all or only offered services that were way out of his ballpark in terms of technological sophistication and cost. Then one day a sales pitch caught his eye. "Seven days to get your business on the Internet," promised the Ohio-based ISP. The deal: for $100 a month, the vendor would host as many pages as Snell needed to list his inventory (100 items to start), provide a shopping cart for secure ordering, and even throw in some extras like bulletin boards and chat rooms. All Snell had to do was program the pages. "I figured, how hard could it be?" says Snell, a graphic designer by training. "I fell for it hook, line, and sinker."

Snell signed up with the ISP in March 1997. Following the company's instructions--including visiting recommended HTML tutorial sites where he learned to build simple Web pages--he had a site up on its server by July. Unfortunately, no one could buy anything from it. Snell says he followed to the letter the ISP's instructions for setting up the shopping cart--and did so over and over--but the thing just wouldn't work. And when Snell tried to pay the ISP for a few hours of consulting, the company said it was too busy. "Their version of 24-hour tech support was that when you sent them an E-mail, it took them 24 hours to get back to you," says Snell, and he was made to feel like a pest every time he called. (The ISP has since gone out of business.)

Snell had a live site, but without on-line order taking it was useless. Rather than promote it, he decided to cut his losses ($500 for five months of hosting) and go looking for a shopping cart that would be easy to set up and to use.

It was about this time that Snell's brother and business partner, Steve, heard about Viaweb (now Yahoo!Store), a site on which businesses can build storefronts using simple templates. Snell signed up for a free trial. Wanting to simplify things as much as possible, he decided to experiment with the inventory from his parents' pet-supply company, which was much simpler than his own. A tutorial walked him through the process. Quickly, he learned how to list items, upload images, and make changes to descriptions and prices. Though he was no stranger to HTML, Snell was relieved to find out that he didn't have to program anything. And most important, the shopping cart worked. Within half an hour--long before the site was complete--Snell successfully placed a test order.

After three days spent mostly at the keyboard, Snell had his mock site: a 50-item catalog, some company information, and a working shopping cart. The ordering mechanism was fairly simple: when people bought items, Snell would receive E-mail or fax alerts including everything but the credit-card information. (To obtain that, he would have had to log in to Viaweb's secure server.) It looked good, but Snell--once bitten--wasn't quite ready to commit. He still had 10 days before his free trial expired, so he decided to examine other options.

The most obvious were on-line malls. Like Viaweb, these sites help merchants create simple storefronts, but most also insist on handling credit-card transactions and want commissions on sales. Snell recalls one mall that didn't even provide individual storefronts but rather mixed together everyone's merchandise like some strange electronic department store. In addition, malls often insist that their tenants make do with URLs that are extensions of the malls' own domain names (www.megamallyourstore). Snell wanted his own domain name, something Viaweb would let him have.

Also in its favor, Viaweb had pages that were simple but looked professional. Snell didn't think that any of the malls he visited had done a better job. Furthermore, several of the mall sites were a bear to navigate. "If it took more than three clicks to order, I didn't want it," he says.

So it was back to Viaweb, where Snell's first monthly payment--$100--bought him enough space to list 50 items. He spent the rest of the summer creating his own site, loading product information and images, customizing pages with Viaweb's tools, and testing the ordering process over and over. His Gun Dog Comics site went live in September 1997, at first under an extension of Viaweb's URL. A few months later, in an attempt to give the site an identity separate from that of his physical store, Snell registered it under the domain name bigbluecomics.com.

Within three weeks Big Blue Comics had its first order--from Singapore. Today the catalog lists up to 1,000 items, and on-line orders constitute 25% to 30% of revenues. Last October, Snell closed one of his retail stores because the Web site was proving more profitable--in part because it had almost no overhead. "E-commerce is going to save the world," Snell says. "And I'm going to be right there when it happens."