He Struck Gold on the Net (Really)

In January 1848, a work crew at John Sutter's mill, near Sacramento, California, came across a few select nuggets of gold. Before long, a half-million prospectors arrived there seeking instant riches. The gold rush was on. Some 153 years later, another gold rush broke out at an old mine called Red Lake, in northwestern Ontario. This time, the fortune hunters wielded geological-modeling software and database mining tools rather than picks and shovels. The big winners were from Australia. And they had never even seen the mine.

Rob McEwen, chairman and CEO of Goldcorp Inc., based in Toronto, had triggered the gold rush by issuing an extraordinary challenge to the world's geologists: We'll show you all of our data on the Red Lake mine online if you tell us where we're likely to find the next 6 million ounces of gold. The prize: a total of $575,000, with a top award of $105,000.

The mining community was flabbergasted. "We've seen very large data sets from government surveys online," says Nick Archibald, managing director of Fractal Graphics, the winning organization from West Perth, Australia. "But for a company to post that information and say, 'Here I am, warts and all,' is quite unusual indeed."

McEwen knew that the contest, which he called the Goldcorp Challenge, entailed big risks. For one thing, it exposed the company to a hostile-takeover bid. But the risks of continuing to do things the old way were even greater. "Mining is one of humanity's oldest industrial pursuits," McEwen says. "This is old old economy. But a mineral discovery is like a technological discovery. There's the same rapid creation of wealth as rising expectations improve profitability. If we could find gold faster, we could really improve the value of the company."

McEwen, a small, soft-spoken man with a neatly trimmed mustache and meticulous tailoring, had one big advantage over his slow-footed competitors: He wasn't a miner, he didn't think like a miner, and he wasn't constrained by a miner's conventional wisdom. As a young man, he went to work for Merrill Lynch, following his father into the investment business. But his father also had a fascination with gold, and McEwen grew up hearing tales of miners, prospectors, and grubstakes at the dinner table. Soon he was bitten by the gold bug too, and he hammered out a template of what he thought a 21st-century gold-mining company should look like. In 1989, he saw his chance. He stepped into a takeover battle as a white knight and emerged as majority owner of an old and underperforming mine in Ontario.

It was hardly a dream come true. The gold market was depressed. The mine's operating costs were high. The miners went on strike. McEwen even got a death threat. But the new owner knew that the mine had potential. "The Red Lake gold district had 2 operating gold mines and 13 former mines that had produced more than 18 million ounces combined," he says. "The mine next door had produced about 10 million ounces. Ours had produced only 3 million."

McEwen believed that the high-grade ore that ran through the neighboring mine was present in parts of the 55,000-acre Red Lake stake -- if only he could find it. His strategy began to take shape at a seminar at MIT in 1999. Company presidents from around the world had come there to learn about advances in information technology. Eventually, the group's attention turned to the Linux operating system and the open-source revolution. "I said, 'Open-source code! That's what I want!' " McEwen recalls.

His reasoning: If he could attract the attention of world-class talent to the problem of finding more gold in Red Lake, just as Linux managed to attract world-class programmers to the cause of better software, he could tap into thousands of minds that he wouldn't normally have access to. He could also speed up exploration and improve his odds of discovery.

At first, Goldcorp's geologists were appalled at the idea of exposing their super-secret data to the world. "This is a very conservative, very private industry," says Dr. James M. Franklin, former chief geoscientist for the Geological Survey of Canada and a judge in the Goldcorp Challenge. "Confidentiality and secrecy about reserves and exploration have been its watchwords. This was a totally unconventional thing to do."

But in March 2000, at an industry meeting, McEwen unveiled the Goldcorp Challenge. The external response was immediate. More than 1,400 scientists, engineers, and geologists from 50 countries downloaded the company's data and started their virtual exploration. When the entries started coming in, the panel of five judges was astonished by the creativity of the submissions. The top winner was a collaboration by two groups in Australia: Fractal Graphics, in West Perth, and Taylor Wall & Associates, in Queensland, which together had developed a powerful 3-D graphical depiction of the mine.