What was Clayton Christopher's worst decision about making tea? The decision to make tea. "We should have outsourced production a lot sooner," says Christopher, whose company, Sweet Leaf Tea, brewed its own for four years. "We would have been three times the size we are now." Perhaps. But such a practical course would have deprived this start-up tale of its folksy moonshiner charm.

The son of a serial entrepreneur, Christopher worked for three years in his father's medical-supply company in Beaumont, Texas, then lit out for Florida, where he captained a charter boat. During his travels, he met a tea merchant who regaled him with tales of the glories of the industry. Tea, decided Christopher, was the business for him. "There are so many channels you can go into: supermarkets, convenience stores, restaurants," he says.

With just $15,000 in savings and a $10,000 investment from his family, Christopher had resourcefulness thrust upon him. Obeying the Texas health department's demand that production take place in an enclosed area, he built an 800-square-foot wooden "tea hut" inside his father's warehouse and installed drains, a water line, and a carbon filter cartridge to extract the chlorine from treated city water. Then he furnished the hut with state-of-the-scrap-yard equipment. A used air-conditioning unit cooled the water inside 50-gallon crawfish pots. King-size pillowcases served as industrial-strength tea bags. Christopher filled the bottles himself -- two at a time -- using a couple of garden hoses. To screw on the bottle caps, he used a Black & Decker drill.

If production was jury-rigged, the product itself was carefully crafted. Christopher spent five months sourcing tea leaves from all over the world. He tried 500 individual teas and another 500 blends before settling on a combination of 60 percent Hunan province and 40 percent Indian. Choosing a brewing process was easier. He used his granny's. "She would brew a really strong tea for not more than five minutes," says Christopher. "Then pour it over ice to stop the brewing process and lock in the flavor."

As for market research: "It was just running around stores and buying different flavors of Snapple, Lipton, AriZona, Nestea," says Christopher. "I knew they spent lots of money on market research and probably knew a thing or two."

Flavor was Sweet Leaf's competitive advantage -- its only one. Christopher, who had taken on his boyhood friend David Smith as a partner, could not afford the fees markets charge for placement on store shelves. They were lucky to get their bottles, dressed in unattractive labels designed on Microsoft Word, onto the bottom shelves of convenience stores. "I'd dust off the cobwebs, set up the bottles, come back a week later, and one had sold," he recalls. A breakthrough finally happened when Christopher persuaded 10 stores to let him place metal washtubs filled with ice and bottles of his tea near the checkout area. People started to notice -- and buy.

Sweet Leaf's first major customer was the Texas-based Market Basket chain. "They didn't see a whole lot of products from Beaumont," says Christopher. Other supermarkets followed, though it would be five years before Christopher and Smith nailed Whole Foods, which put Sweet Leaf on the map. "We called Whole Foods for about a year and sent samples but never heard from them," says Christopher. "Finally, David got a postcard. You could tell the buyers had stacks of them in their drawers. It said, 'Thank you. We are...' then there was a line, and someone had written 'NOT...interested in your product at this time.' David was like, 'Look! We got something from them.' We laughed and slapped high-fives. At least they knew we existed."

The partners hired college students to brew and bottle while they made deliveries in a refrigerated Otis Spunkmeyer van with 280,000 miles on it. They spent weekends in grocery stores, talking tea to shoppers and passing out samples. "At least 90 percent of our marketing went to free samples," says Christopher. "You've got to get the product past people's lips."

But the partners' homegrown plant couldn't manufacture a shelf-stable product, which prevented Sweet Leaf from contracting with a beverage distributor. Instead, the two were forced to work through egg and milk distributors, who know their way around perishable goods but not around competitive, branded products. In 2002, the pair finally shut down their plant and signed up with a bottling company that pasteurized the tea, extending its shelf life.

Though Sweet Leaf is national now, some things haven't changed. Free samples -- at live music events as well as stores -- remain the chief mode of marketing. The company spends about 8 percent of sales on freebies. And the Sweet Leaf staff still creates all the flavors. "Lots of beverage companies outsource their R&D," says Christopher. "But the flavor is the most important part of the equation. No one cares more about your flavor than you do."