When Mark Hoffman was pushed out as CEO of Sybase (SY ) in the summer of 1996, his first thought was to retire to his 20-acre horse ranch in the shadow of the Bay Area's Mt. Diablo. He had co-founded the company and built it into the world's then-second-largest maker of relational database software. Yet he had also made a crucial mistake -- producing a new version of Sybase's database in the mid-1990s that didn't sync well with applications from the likes of Oracle Corp. (ORCL ) and Germany's SAP (SAP ). That hurt sales. When Sybase ran into trouble, Hoffman was shoved out the door.
Just a couple of weeks later, though, his hopes were reborn. Jeffrey T. Webber, a venture capitalist pal of Hoffman's, had suggested he poke around at a tiny company, DistriVision in suburban Concord, Calif., that was creating electronic product catalogs on CD-ROMs. The founders didn't have much polish. Tom Gonzales Sr., who wore jeans and cowboy boots, had trained ex-cons how to drive 18-wheelers before the PC bug bit and he started setting up computer systems for mom-and-pop businesses. His son, Thomas Jr., was a quiet, stay-up-all-night programmer.
Hoffman saw huge potential. The world had discovered the Internet. While most people were fascinated with consumer Web sites such as Amazon (AMZN ) and Yahoo! (YHOO ), Hoffman foresaw a time when the Net could be harnessed to automate the dealings between companies, saving billions of dollars in costs.
In DistriVision, he glimpsed the glimmerings of a way to do that. Rather than deliver electronic catalogs on disks, DistriVision could put them on the Internet, where anybody could easily get to them. Hoffman volunteered to run the company -- soon changing its name to Commerce One (CMRC ) and moving to bigger digs in the California towns of Walnut Creek and then Pleasanton.
When dot-coms posed a grave threat to old-line businesses, Hoffman came off as a trusted friend
No flashy leader, Hoffman is 5'8", with a placid face and a voice that rarely rises above a whisper. But at a time when dot-coms posed a grave threat to old-line businesses, he came off as a trusted friend of the Establishment. Alexis de Raadt-St. James, later an executive at Commerce One, recalls meeting Hoffman at a technology conference in Monaco in 1999. She was Shell International's e-commerce chief at the time. At a cocktail party, a then-famous dot-com CEO told her contemptuously that he "'wouldn't work with Shell if it was the last company on earth.'"
Hoffman was a stark contrast. He asked her detailed questions about Shell's business processes. "I could see this was fascinating for him. He could see what would come next," says de Raadt-St. James.
The Season of Hype
What came next defied all reason. It also planted the seeds for Commerce One's eventually fall from grace. While some dot-coms already were starting to fail in 1999, it was the business-to-business (B2B) e-commerce companies' turn to rocket toward the stars. Analysts confidently predicted that by 2005, $6 trillion of commerce between businesses would be handled via the Net. Large tech companies such as Microsoft (MSFT ), Oracle, and IBM (IBM ) weren't even in the game. The leaders were Ariba Technologies (ARBA ) and Commerce One. And they made the most of it.
Commerce One became a marketing machine. Hoffman and then-strategy chief Chuck Donchess spoke at practically every B2B conference -- and there were dozens. The company held its own conferences, too, called eLink. All it had to do to get in the news was issue a press release, and a dozen reporters would call. "It was like throwing cows into a lake full of piranhas," says Andrew McCarthy, who was then director of PR.
The rivalry with Ariba was a gift. With these two companies running neck and neck, every feat of one-upmanship made good copy. Commerce One used the competition to whip its troops into a frenzy. Then-sales chief Mark Biestman, whose nickname was The Biest, passed out "Aribanator" T-shirts--with pictures on the front of Arnold Schwarznegger as The Terminator. One day, dressed as General Patton, he drove a Harley-Davidson motorcycle into a hotel during a sales training event.
Heeding the Get Big Fast mantra, Commerce One didn't even set up a beta-testing program for new software
No behavior seemed too outrageous. A few weeks before Commerce One's July 1, 1999, initial public offering, the executives gathered at MoMo's, a trendy San Francisco restaurant that some jokingly called Commerce Two because they spent so much time there. The occasion: a deal to sell software to Japanese telecom giant NTT (NTT ). Webber, the venture capitalist, who was then on the board of directors, remembers peeling off a J Wine Company sweatshirt and a Commerce One polo shirt, and, bare-chested, presenting them ceremoniously to the Japanese. Everybody hooted with laughter. "It was a fun culture," he says.
But the go-go attitude opened the door to mistakes that didn't show up until later. In the glow of easy victories, Hoffman didn't keep a close enough watch on basic operations, according to several former managers. Commerce One's engineers and marketers were frequently at loggerheads. Heeding the Get Big Fast mantra of the moment, the company didn't even set up a beta-testing program for new software releases. The result: an array of products that didn't work well together. Customers suffered, and ultimately it spoiled Commerce One's reputation.
The Road to Rapid Riches
During its heyday, everything Commerce One touched seemed to turn to dough. Its stock rose by more than 1,000% by the end of 1999, then peaked on Mar. 9, 2000 with a $21.5 billion market cap. The key to its success was the emergence of a new phenomenon, the electronic marketplace, which also led to Commerce One's undoing.
This was Hoffman's baby. Back in 1997, he quietly started building software for e-marketplaces, Web sites where sellers and buyers would come together to trade everything from paper clips to raw steel. Commerce One's breakthrough came in the Old Economy bastion of Detroit. In October of 1999, General Motors (GM ) procurement chief Harold R. Kutner staged a bake-off between Commerce One and software giant Oracle for the right to power GM's e-marketplace. The carmaker set harsh terms. It wanted stock in the tech companies, and it wouldn't pay for software or services. Oracle refused to play by those rules.
Customers were competing with one another to land Commerce One as their supplier
At first, Hoffman also recoiled. But while taking a shower at home in California a day after hearing GM's ultimatum, he realized that running its trading exchange could set Commerce One up for huge transaction fees. He decided it was worth the gamble, so he agreed to negotiate on those terms. After around-the-clock talks in Detroit on Oct. 29, Hoffman and Donchess ran out of clean clothes and bought University of Michigan duds at a hotel shop -- hoping that would curry a little favor with the GMers. The next day, they struck a deal.
The GM deal set off a business bonanza that made Commerce One the fastest-growing Nasdaq company -- ever. It signed up dozens of e-marketplaces, from aerospace to wood products. And at the peak, customers even competed with one another to land it as their supplier. After logging just $33.6 million in sales in 1999, Commerce One racked up $401 million in 2000, and was on a pace to double that in 2001.
Even while Commerce One was riding high, Hoffman fretted. At a brainstorming session at the Walnut Creek Marriott in January, 2000, he worried about a new alliance between Ariba, IBM, and supply-chain software maker i2 Technologies (ITWO ). Commerce One needed a strong partner that would lend it credence in giant corporations that were more comfortable buying software from the likes of IBM. SAP, the king of corporate software, looked like the best choice.
Hoffman called SAP Co-CEO Hasso Plattner and got turned down. But a month later, Plattner reconsidered. SAP was getting written off as a has-been for falling behind on the Internet, and an alliance with Commerce One would put that issue to rest.
The first face-to-face meeting almost didn't happen. On his way to the W Hotel in San Francisco, Plattner hit a speed bump in his Aston Martin and scraped off his engine's oil pan. He was 45 minutes late, but Hoffman waited. By June, the two companies had a deal. They would combine their engineering teams to deliver a new set of products and then unleash their sales teams on Corporate America. No two technology companies had ever tried to combine forces so thoroughly -- short of merging. The coupling brought Commerce One instant credibility and nearly $500 million in cash.
Nothing captures the jittery energy of that era better than the last days leading up to the deal being struck. SAP and Commerce One people met, and regrouped, and met again in Las Vegas hotels and conference rooms leading up to SAP's big user conference. They finally finished at 8 a.m., just before Plattner and Hoffman were to appear on-stage before an audience of thousands.
By the end of 2000, rank-and-file employees were holding stock options worth hundreds of thousands
Gary Fromer, then chief strategy officer for SAP's e-marketplace subsidiary, recalls sitting in the audience when the two CEOs strode into the spotlights. And suddenly he couldn't stand it. "I had spent so much time and effort on it that I couldn't watch. I got up and walked out," he says. Just as he got to the exit, he saw that Rob Tartkoff, his counterpart at Commerce One, was also heading out. They went off to the blackjack tables together -- and won about $500 apiece.
By the end of 2000, Commerce One's fortunes were soaring. The company boasted 567 customers and 157 e-marketplaces. Everybody there got a taste of riches. Rank-and-file employees saw their stock options grow to be worth hundreds of thousands of dollars. Top salespeople became ridiculously wealthy. And customers and partners shared in the bounty. At the peak, Hoffman hosted 2,800 people at a conference in the swanky Bellagio Hotel in Las Vegas. He watched in the Drink nightclub while hundreds of his guests danced wildly to the R&B riffs of Jim Belushi and his Sacred Heart Band.
Downhill from Here
The problem was, it didn't get any better than that. In the first quarter of 2001, demand began to taper off. Revenues of $170 million were down 10% from the previous quarter. At Commerce One's eLink event in Berlin, starting on Feb. 19, only 1,000 guests showed up. Berlin's Brandenburg Gate was wrapped in cloth for repairs, and to some Commerce One people, it seemed the same would soon be true for them.
What went wrong? A bunch of things. Some of the early e-marketplaces took many months to get going, partly because of problems with the technology but mainly because the industry consortiums that formed the e-marketplaces had trouble getting coordinated. Those delays put a damper on hopes for quick and easy riches.
Then the slow economy put the brakes on all kinds of technology spending. "People were buying vision in 2000. In 2001, they were looking for return on investment," says Robert M. Kimmitt, then Commerce One's president and now an executive vice-president at AOL Time Warner (AOL ). When the new products from Commerce One and SAP were ready to conquer the world in early 2001, the world had changed. E-marketplace software was no longer in demand.
The team eventually retooled the products to handle private e-marketplaces, where individual corporations would interact with their suppliers. But that fell flat, too. And so did Commerce One's revenues, dropping to $101.25 million in the second quarter of 2001 -- roughly half of their peak.
Commerce One's bet on linking with SAP left it uncompetitive as a solo act
By then, Commerce One's fate was totally entangled with SAP. The U.S. company had spent a year focusing its 400 engineers on combining its technology with SAP's e-commerce applications. When hard times arrived, Ariba was able to strip away everything else and fall back on its e-procurement business, which is desktop software used by corporate employees for ordering supplies via the Net.
But Commerce One had bet everything on the SAP relationship and had let its e-procurement software slip. It wasn't competitive anymore as a stand-alone company. So when Plattner proposed merger talks during an August 10, 2001, lunch at SAP's Silicon Valley offices, Hoffman put aside his dreams of building an independent software powerhouse and said yes.
Despite all the signals that the glory days were over, Hoffman had a hard time accepting it. Commerce One had more than 3,000 employees through most of 2001. It still promoted itself as the worldwide leader in e-marketplace software long after the market had gone sour. Hoffman confesses that he kept hoping the economy would turn around.
Instead, he kept building for a brighter future. Dennis Jones, the longtime chief information officer of Federal Express, came on as Commerce One's chief operations officer in April of 2001 to provide big-company business processes. Donchess resigned for personal reasons, and Hoffman replaced him with de Raadt-St. James. She had just finished an MIT masters degree. Her thesis topic: managing expansive growth.
So just when Commerce One needed to shrink and refocus smartly, its new executives were best suited for a big company that was getting even bigger. "It was like bringing in a podiatrist when what you needed was a brain surgeon," says Roy Satterthwaite, a former marketing vice-president.
Undoing the SAP relationship practically ruined Commerce One. The deal fell through a couple of weeks after the September 11 terror attacks. Plattner, who had forged the alliance with Hoffman and was loath to give up on it, got a back-channel call from a board member. "They pre-preempted it. They asked me not to spend any more money on Commerce One," says Plattner.
Hoffman already had a gut feeling that something was amiss. The intensity of merger talks had dropped off precipitously after the attacks. Werner Brandt, SAP's chief financial officer, didn't mince words when he called Hoffman in late September. "'It's not going to work out, given all the stuff that's going on,'" Hoffman recalls Brandt saying. Looking back, Hoffman wishes he had handled the alliance differently from the start. "I wouldn't let myself so closely align" with another company, he says. "It has been very difficult to extract."
Indeed, breaking up was extremely hard to do. The companies' products were thoroughly integrated with one another, and they were pushing a single e-marketplace package. Because most major corporations already had relationships with SAP, many of the new deals were written as addenda to existing SAP contracts. Now, though they maintained a marketing partnership, most of their relationship had to be unwound.
At the same time, SAP started warning corporations that they had better play it safe and buy technology from big, stable suppliers -- undercutting Commerce One. Hoffman had learned the hard way to avoid intertangling alliances.
The Internal Split
At a time when Hoffman needed a unified team behind him, his lieutenants were at war over the company's soul. It was the newbies vs. the old guard. When Jones had proposed Commerce One's first-ever beta-testing program in the summer of 2001, he caught flack from managers who believed it would slow things down when new products needed to be launched quickly.
Some Commerce One veterans thought de Raadt-St. James wasn't qualified to set strategy for a software company. They privately mocked her frequent references to "S-curves" -- a way of tracking growth in a booming market -- as laughable pedantry. "We didn't mix well from a cultural and style point of view," says Mike Micucci, who then ran marketing and quit when he lost a power struggle with her.
De Raadt-St. James acknowledges that she met plenty of resistance. "I caused a lot of tension around running a company in a very structured way," she says, recalling that people came into her office, "angry, crying, saying they were leaving." That was something she urged them to do if they felt they couldn't adapt to a new way of operating.
"I have to go on my gut. I don't like our chances"
Indeed, many of Commerce One's faithful were abandoning ship. Strategic marketer Mark Nittler and merger expert Robert Tarkoff left before the end of 2001. At a Jan. 7 dinner at Moose's restaurant in San Francisco, Hoffman asked Biestman to hold on for six months, by which time he hoped the company would begin to show signs of life. "I said, 'I have to go on my gut. I don't like our chances,'" recalls Biestman, who's now CEO of SEVEN Networks, a telecom startup. "I told him I thought we had lost our way as a management team." Three months later, The Biest was gone.
Just when Commerce One hit the skids, co-founder Thomas Gonzales Jr. died of a rare form of cancer. Last Dec. 13, more than 600 friends and colleagues gathered for an evening memorial service at the Blackhawk Auto Museum in the hills east of San Francisco. Like the treasures of a pharaoh, Gonzales' prize possessions, F1 McClaren and Ferrari F50 sports cars, were placed on display -- as were drawings of the dream house he never got to build.
At the end, people lingered in the parking lot in the dark. For some, this was the ultimate downer. Gonzales grew up poor in a cabin in California's Sierras. He had worked hard. He was a team player -- always ready to stay up all night to finish a project or fly to any city to show the company's newest technology to potential customers. And thanks to the Internet gold rush, he had attained wealth beyond his dreams. Now he had died way too young. "When Thomas died, something died within the company. The spirit went out," says Biestman.
Indeed, the Blackhawk farewell is a fitting epitaph for a whole generation of Internet startups. They came. They soared. And, like Commerce One, they're in danger of fading away without fulfilling their promise.
By Steve Hamm in New York