Over the course of a few months, Benchmark Capital had explored the possibility of dot-com ventures with Goldman Sachs and some Fortune 500 partners, but they had dithered endlessly. Bob Nakasone, chief executive of Toys "R" Us, moved with dispatch. At the end of a two-hour meeting on a Wednesday, Nakasone said, "I want to be in business with you guys. I’ll fax you a term sheet tomorrow," which he did. Bruce Dunlevie, a Benchmark partner, worked on the draft over the weekend. By Monday, Toys "R" Us and Benchmark had agreed on the deal. Two days later, a public announcement was made.

That Toys "R" Us would commit so quickly suggested that this was an organization that had the fast clock rate of a Silicon Valley company. It was not so simple, however. The following week, when the Toys "R" Us lawyers sent the final term sheet, much had been changed, reflecting the desire of Toys "R" Us’s finance and accounting organizations to use the new entity for their own tax purposes. In Benchmark’s eyes, this was a term sheet that had hair and complexity all over it.

At Benchmark’s 1999 annual meeting, in early June, Dunlevie already seemed hesitant. Dunlevie, the partner who went on the board of the Toysrus.com venture told guests that it was too early to say what the outcome of the experiment was likely to be. "It will be a very positive, signal event of the dot-comming of America, or it will be a large disaster," he said.

Dunlevie did defend the decision to take on revamping a Fortune 500. The skills that would be required to do Toysrus.com right were highly similar to what Benchmark did: Recruit the team, craft a strategy that made sense, and execute it.

"Does an eToys first-mover advantage and the looming Amazon.com threat make this a bad idea?" he asked. He had reason to be optimistic: "This is very different than Amazon books. The supply chain in toys is incredibly complicated. There are thousands of sourcing relationships. You don’t just get to drop-ship them in Memphis. You’ve got to buy unit quantity—250,000—in China, get them on a boat and to the U.S., and distribute them to your stores, or, in the case of the dot-com business, get them into your Memphis warehouse."

Toysrus.com was one dot-com that would enjoy superior buying power right from the outset. No other company in the world bought more toys and games than Toys "R" Us. Most important, the dot-com business would have access to the parent’s acquisition power—and the ability to get the items that were in short supply every Christmas season. Other toy retailers would have to beg for what they thought would be a fair allocation, but Toys "R" Us could go to the manufacturer and say, I’ll take every single one you make and COD. In the book business, if a product is in short supply, the publisher simply goes back to the presses, but in the toy business, the hot toys of a season cannot be produced and delivered fast enough.

Dunlevie was excited about the prospect of using customer "touches" in the legacy bricks-and-mortar business to pull people to the Web. Plans were being drafted to place greeters in the stores in neighborhoods where there was evidence of high Web usage. The greeters would offer to escort customers to a Web terminal set up in the lobby to show off the dot-com site and offer inducements to use the on-line business in the future.

"This is a very different mind-set from that at most companies, which are trying to protect what they’ve got," Dunlevie said at the Benchmark meeting. "You have to decide you’re going to eat your own business, as opposed to having eToys or Amazon or somebody else doing it for you."

"You’ve got to do it the right way," Dunlevie added. "You’ve got to set up a separate company. The management team and the board of directors have to be fully independent and autonomous." The "right way" also meant the dot-com business had to offer the financial upside of stock options to employees and have unlimited access to use of the legacy business assets. The parent had to be ready for "100% self-cannibalization," he said.

Toysrus.com seemed well-positioned to be a huge success. But perhaps thinking silently of those niggling legal issues that had not been fully resolved, Dunlevie ended his otherwise upbeat presentation on a jarringly ambiguous note: "Time will tell."  

The telling began almost immediately, and it was not encouraging. Less than 10 days later, Nakasone met with the man who had overseen Toys "R" Us’s earlier on-line efforts. The man said that Toysrus.com’s new president, Bob Moog—who had been endorsed by Benchmark partners—was a disaster. Nakasone hopped on a plane for California, asked Moog to dinner, and said he’d lost confidence in him. The next day, Nakasone fired him.

Later that day, Nakasone faxed Dunlevie a revised equity budget for the dot-com subsidiary. Instead of reserving 20% of the company’s equity for employees, the figure would be 10.5%. Nakasone said Salomon Brothers assured him that 10.5% would be sufficient for hiring 300 people. Where Salomon had looked, Nakasone did not say, but Dunlevie felt that 10.5% was barely adequate for a single CEO, let alone 300 employees.

Dunlevie interpreted the fax as reflecting Nakasone’s realization that, under the original equity budget, the executives of the new entity would likely have become wealthier than everyone else in the old business. The disagreement on equity exposed the difficulty of having two very different compensation models placed side by side.

Then Dunlevie met with a Toys "R" Us executive who showed him a business plan for the dot-com subsidiary that called for $25 million in revenue that year.

Dunlevie was puzzled. "I thought it was going to be $100 million."

"For that, we would need access to the hot products," the man replied. "If we could just get our hands on Harley Barbie, Harley Ken, and Sega Dreamcast, I could easily do $100 million."

Dunlevie said: "I talked to Nakasone on this very issue on Friday morning, and he said you can get whatever you want."

"Well, it’s not really true," Dunlevie was told. The dot-com subsidiary would receive only a small portion of those products.

Dunlevie returned home and attached himself to the telephone. In a conference call with Nakasone and senior Toys "R" Us managers, Dunlevie brought up the most immediate sore point, the internal pricing of merchandise. The latest document from Toys "R" Us attorneys said the parent would tack a 5% to 15% charge on all products it provided the dot-com business.

This was the epitome of how not to do the dot-comming of America. Dunlevie was provoked to do something he rarely did: He lost his temper. He said he was going to join his family for dinner. "You guys can continue if you want to," he said. He hung up.

Three minutes later, Nakasone called back. "C’mon, partner, let’s get back at this."

"Bob," Dunlevie said, "this deal gets worse the closer I get to it. None of what you represented is true. The organizational dis-preference for trying to accomplish this is obvious." Nakasone unilaterally struck the merchandise charge, saying, "Everything you get, you get at cost."

They then turned to another touchy subject, the allocation of hot products. To Dunlevie, this was the touchstone for the deal.

"When we talked about this eight weeks ago," he reminded Nakasone, "you said dot-com got everything it wanted. So what I’m telling you now is, I want all of the Sega Dreamcasts in dot-com. That’s $40 million of business. The product is going to be in shortage. There were only going to be 300,000 shipped to the U.S. Of those, 200,000 are coming to Toys "R" Us. I want all of them. And I want the Furbies, Harley Barbie, and Harley Ken."

"It’s not that simple," Nakasone said. "We’ve already committed to catalogs; store managers are already taking presales. What do you want us to do? Go into stores and pull things off the shelves?"

"I’m not talking about that," Dunlevie said. "I’m saying that, with things you know ahead of time will be in limited quantity, Toysrus.com gets them all." [The Harley Barbie exemplified Dunlevie’s problem. The new doll was produced in a three-way deal between Harley-Davidson, Mattel, and Toys "R" Us, so the only distributor in the world would be Toys "R" Us. Of 134,000 pieces, the dot-com business was going to get only 17,500.]

They spent an hour on the issue. In the end, they agreed to put the matter on hold because they were not speaking the same language.

"Look," Dunlevie said, "the goal of the dot-com business is to cannibalize the stores."

Nakasone was silent for a long time before speaking. "That’s not the goal at all. The goal is to complement."

"OK," Dunlevie said, "I’ll correct my language. The end result is going to be the cannibalization of stores. You’re going to go from 800 stores to 200 in the next three or four or five years as a result of the Web."

Nakasone did not agree. Dunlevie concluded that he wanted everything—the existing network of stores and a thriving on-line business. That was a natural-enough inclination but, Dunlevie felt, a delusional expectation.

Not long afterward, the partnership collapsed. Nakasone was forced out of the company.

[The stock price at Toys "R" Us is off almost 50% from last year’s high, reached during the early enthusiasm for the deal with Benchmark. Toys "R" Us, which has more than 300 times the revenue of eToys, has barely twice the stock-market value.]

A week after the split with Toys "R" Us, Benchmark announced a deal with Nordstrom to launch a new on-line site, nordstromshoes.com. Benchmark contributed $15 million, one of its largest investments ever.

This time, negotiations had gone swiftly and smoothly, helped by the 36-year-old executive whom the parent company had placed on point and who would become the new unit’s CEO. An avid skier and climber, he got along well with the Benchmark partners when he came down from Seattle to visit. He declared that Nordstrom had chosen Benchmark not only for its e-commerce expertise but also for help in developing a corporate culture at Nordstrom "just as dynamic and hard-driven as a Silicon Valley start-up."

At the onset, Toys "R" Us had said all the right things, too, but in Nordstrom’s case the details agreed upon between the two partners had remained intact, free of CFO and attorney meddling on the part of the larger partner. The new subsidiary would have its own management team and its own board of directors, and the sourcing arrangements with the parent were clean, without any surprise gotchas.

Why had Nordstrom executives been able to avoid falling prey to the fear that wealth generated by the dot-com side would create a rift between haves and have-nots? Why was Nordstrom able to follow through, organizationally, when Toys "R" Us had not? Perhaps a clue is found in the selection of that young executive appointed to head the dot-com. His name is Dan Nordstrom. He was one of six co-presidents who shared the surname, a youthful group of brothers and cousins, all in their thirties. Although Nordstrom’s shares were publicly traded and the CEO was not a Nordstrom, the family owned 35% of the company, wielding power that could obliterate organizational inertia. In this case, family ties served to speed internal change.

Adapted from eBoys: The First Inside Account of Venture Capitalists at Work by Randall E. Stross. Copyright ©2000, Randall E. Stross. Reprinted with permission of Crown Publishing, a division of Random House, Inc.