Running Money, page 18
In fact, maybe Metcalfe’s Law is the formula for Doug Engelbart’s scaling of human knowledge, just as Watt’s steam engine scaled human power.
And Xerox PARC? I’ve been there a few times. The first was in the early 1990s, when Xerox management was hoping to impress Wall Street with a modernization kick. Xerox could commercialize projects, blah, blah, blah. I asked about the beanbag chairs and got quickly sniffed at: “That was the old Xerox PARC.” Now they were serious.
In 1999, I was asked to meet an entrepreneur at Xerox PARC who was trying to commercialize a virtual whiteboard collaborative thingy. I never could really understand its uses. On the way to his office, we walked past this huge room filled with goofy scientists sipping lattes, each resting comfortably in his or her own red, green or blue beanbag chair. It’s good to know that some things never change.
Metcalfe helped local networks scale—and scale they did. Every department inside a company was able to share files and printers, replacing sneakernet. Dealing with the outside world was another story.
It was the personal computer boom of the 1980s that dragged LANs into the mainstream. But they were not all the same flavor. There was Ethernet, of course, but Novell had a networking standard in Netware, IBM had Token Ring, UNIX computers used uucp, etc. It was a mess.
Around 1980, Xerox PARC gave Stanford University a bunch of Alto workstations and some of their new Ethernet networking cards. Like any large university, Stanford had a huge collection of computers: mainframes, minicomputers and even some homegrown Motorola microprocessor-based machines built at Stanford by grad student Andy Bechtolsheim, who would later use them to found Sun Microsystems. Plus, all these computers were scattered among different schools doing research—the med school, the engineering department, the business school, etc.
Len Bosack and Sandy Lerner get credit for inventing routers and then founding Cisco to sell them. These routers were basically miniversions of the store-and-forward IMP that Kleinrock and others used for ARPANET, except companies and universities could use them to hook up all their scattered departments and LANs, even across the country.
Still, 80% plus of all network traffic was on these LANs. Sure, the Internet had existed since Kleinrock’s L-O crash, but no one outside of academics used it much.
In the early 1990s, as these networks proliferated, another pressure point was building, another barrier to be broken, unleashing all those users to the outside world. This would turn out to be the big one to surf.
Let me take you back to a story near the beginning of all this.
Clark’s Outpost
WOODSIDE, CALIFORNIA—SPRING 1994
“You’re such an asshole. John Doerr doesn’t know everything,” Bob Harris launched almost immediately.
“Yeah, well, that’s what I’m doing,” Jim Clark said.
“Well, you’re being stupid.”
“Me stupid? You thought I was working on MOTIF.”
“Motif, Mosaic, who gives a shit. We can help more than some know-nothing VC,” Bob screamed.
“Shit,” I whispered to my wife, Nancy. “I think we are a little late. About three highballs too late.”
Late indeed.
I got to know Jim Clark back when he was chairman of the computer company Silicon Graphics and I was still with Morgan Stanley. Frank Quattrone introduced us at one of the many road show lunches. Silicon Graphics was an investment banker’s dream, always raising money in the stock market to fund one project or another. Clark was the technical founder, and he always seemed to be restless. I was writing about some new uses of technology, and Clark was intrigued. Back in 1991, he was pitching a new concept known as the teleputer, a computer, to be made by Silicon Graphics naturally, that would be networked with every other computer in the world and provide information, entertainment and everything else people might want. Pretty cool for 1991.
But as smart as Clark was in technology, he was a business dope. In fact, when he started Silicon Graphics, he sold two-thirds of the company to venture capitalists Dick Kramlich of New Enterprise Associates and Glenn Mueller of Mayfield for $600,000. To add insult to injury, these two VCs stayed on Silicon Graphics’ board of directors and turned down most of Clark’s requests. He asked for funding for new projects, like the teleputer. But he also asked for rich option and pay packages, probably in an attempt to get his fair share to make up for his getting screwed in the seed round. The board fought him on virtually everything he proposed, which from what I could gather, pissed Jim Clark off to no end.
I started working with Bob Harris in 1991. I thought Bob was the best investment banker in Silicon Valley. The list of companies he had taken public was staggering, from Microsoft to Sun Microsystems, but the company he was closest to was Silicon Graphics. He was their go-to banker. The board called in Bob Harris and asked for his honest opinion, something in short supply from fee-hungry investment bankers. I would always find myself sitting in the passenger seat of Bob’s green BMW 540, scooting around the Valley, looking for deals, and would get dragged into a few too many meetings at Silicon Graphics.
The last few meetings were the most fun. Clark was fed up. He wanted out. TimeWarner had picked up on his teleputer concept and wanted to roll it out in Orlando, Florida, as something called the Full Service Network. Clark desperately wanted to be involved, but others at SGI cut Clark out of all the discussions. The Full Service Network flopped miserably and probably would have with or without Clark involved. Clark’s last straw was to hit the ejection button from Silicon Graphics.
Bob gave Clark advice every step of the way, probably jeopardizing his working relationship with the Silicon Graphics board.
So Bob Harris and Jim Clark went way back. Long enough for Harris to be able to call Clark an asshole in a public restaurant.
“Jim, this is my wife, Nancy,” I said.
“Hi, Nancy, nice to meet you. Thank god you guys are here,” Jim said as he rolled his eyes. “Let’s eat.”
“Nancy, this is Jim’s wife, Nancy Rutter,” I said.
“Hey, another Nancy. You guys showed up just in time, it was starting to get ugly,” I said, trying to change the subject away from Harris yelling at Clark.
“Nancy Rutter works at Forbes ASAP with Rich Karlgaard. I think she even edited some of my columns over the years.”
Rich Karlgaard, now the publisher of Forbes magazine, told me the story about a very pretty Nancy Rutter going to interview venture capitalist Dick Kramlich for Upside magazine. By chance, she met Jim Clark in the lobby, and a week later, after some follow-up, Nancy Rutter was seen driving around in a Mercedes 450 SL.
“Why is Bob yelling at Jim?” my wife asked.
“I don’t know,” I said.
“You stupid son of a bitch,” Bob continued.
“He needs to be yelled at,” Nancy Rutter explained.
“Were you guys just on your boat somewhere?” I asked, again trying to steer the conversation to calmer waters.
“We just got back from Fiji. Didn’t Jim call you from there?” Nancy Rutter answered.
“He called me from the boat while I was at some conference, ship to shore or satellite phone, sounded great,” I said.
“Me stupid?” Clark screamed back to Bob.
“Muzak. Munsters, what the hell,” Bob mumbled.
I thought it was time for me to break it up. “So, Jim. I’m an Illinois alum, and Mosaic looks pretty cool as a front end to a lot of things. Tell me more about the deal, and I’ll get the barking dog here to calm down for a few seconds.”
“I’ve already put in a couple of million. The deal now is $5 million for 20%. I’ve got all sorts of people bugging me. Kramlich calls every day. Mueller whines on the phone not to be left behind,” Clark said in a complaining tone.
“Well, we’re interested. Our fund just did a first closing. By the way, we’re still waiting for your money. But we are all set to invest,” I said.
“I think I’ve decided John Doerr and Kleiner is going to do the whole investment themselves,” Clark said matter-of-factly.
“John Doerr? Jesus Christ. He’s not the only one that knows anything.” Bob just wouldn’t let up. And we hadn’t even ordered our food yet. “You’re going to need a lot of help. We know this space.”
“Remember William Morris?” Harris asked. We had set Clark up with an entire department of William Morris Agency to look into where technology and entertainment crossed.
“That was OK.”
“Andy and I had breakfast with Brandon Tartikoff and those William Morris guys. Tartikoff said he enjoyed meeting with you,” Harris said without yelling.
“I suppose. John Doerr sort of insists on doing this one himself,” Clark blurted out.
“What?” Bob screeched.
“Yeah, he wants the whole $5 mill and nobody else. He’s pretty adamant,” Clark said.
“I think there’s lots of ways to go,” I said, jumping in.
“What do you mean, Andy?” Jim asked as he turned his back on Bob.
“Clark’s Outpost,” I said.
“What the hell is that?”
“I’ve been doing a lot of work on how this whole interactive media is going to be paid for. Mass advertising is awful. TV charges a buck for 1,000 impressions. That won’t work on PCs. Once you can measure response rates, that model will die.”
“I just came back from New York. I met with a bunch of magazines and told them we’d like to charge them a nickel or a dime for each page that people on the Internet view with Mosaic,” Clark said.
“The way to protect the software is to make it the front end to something else. I’d sell people stuff, that’s what Clark’s Outpost is,” I offered.
“I don’t get it.”
“It’s direct advertising. Junk mail guys pay $400 per thousand mailings, hoping to get a 2% hit rate. If they can clear $20 on what they are selling, they break even. Everything over 2%, and they wipe up.”
“So?”
“So, just put your face on the main Mosaic page.”
“And?”
“And you’ll get 20% hit rates. Everyone will click on your face to head into Clark’s Outpost. You can sell them anything you want. PCs, TVs, video games, even books, I suppose.”
“Sounds messy. I like software.”
“I like software too, the margins are great, but it’s impossible to protect. Microsoft has a business because they tie Windows to real hardware. Tough to pull that off again. But you can leverage yours with that direct marketing model, and you’ll own the Internet.”
“And drop that Doerr guy,” Harris slurred.
Kleiner Perkins did the Mosaic deal by themselves, shutting my ass out. Why am I telling you this sad story? Because Mosaic broke the 80-20 barrier.
Marc Andreessen was working at the National Center for Supercomputing Applications at the University of Illinois and was sick of typing in command line prompts to get what he wanted from the Internet. So he and a few other folks devised a browser—with pages of links to other pages. Actually, it was what Doug Engelbart had envisioned and even crudely demo’ed in 1968. Andreessen made it hum by allowing his browser to pull multiple packets at a time from the Internet rather than one at a time. The free code spread like crazy. Users were no longer stuck on the LAN—they could browse the outside world. Clark met with Andreessen and hired him and his buddies for a company they named Mosaic Communications.
It was soon renamed Netscape when the University of Illinois sued over the name. I found out later that indeed John Doerr didn’t want anyone else involved. Hey, all’s fair. They had deep pockets and could set the rules. Kramlich and Mueller continued to hound Clark to be let in the deal, especially as Netscape began to be touted as the next Microsoft.
Our little venture fund did close, without Clark’s money, and we funded a bunch of companies that did well because Netscape browsers were spreading faster than wildfires in a Texas prairie.
A few months later, Glenn Mueller took his own life on his own boat while down in Mexico. Mueller was not only one of the top five most successful venture capitalists in Silicon Valley, his wife, Nancy Mueller, had a hugely successful catering business. These two were the toasts of the town. Why Glenn Mueller took his own life is still a mystery, although there were rumors running around that he had bouts of depression or other emotional problems. Who knows? But Jim Clark took it hard. He blamed himself for cutting Mueller out of the deal. Clark gave the eulogy at the memorial service for Mueller. Bob Harris was there and told me that Clark broke down crying several times.
Over the next year, things moved fast at Netscape. Jim Barksdale, a former FedEx and AT&T Wireless executive, was brought in as CEO. Clark was “elevated” to chairman and, I think, had very little to do with managing the company. But he did own a big chunk of it. The business model that evolved for Netscape had nothing to do with charging a nickel or a dime or even a penny for each page viewed. Instead, they just gave away the browser to anyone who could go to their Web site and download it. Millions did.
Then Barksdale went around to corporations and told them that they were using thousands of copies of Netscape’s browser software and that it was free only for consumers, not corporate users, so please pay up. Quite ingenious.
Revenues grew, and Netscape sucked up every available programmer in the Valley to create new versions of the browser and server software to spit out the Web pages to millions of users.
On the day of Netscape’s IPO in August 1985, the share price popped from $28 to $54 in the first few minutes of trading, valuing the company at $2 billion. I walked into Bob Harris’s office.
“Did you see where it closed?” I asked.
“Yeah, don’t remind me.”
“What, that we were three highballs away from investing in Netscape?”
“Yeah, that. And Clark is going to be an even more insufferable prick.”
“Gotta get over it, Bob,” I said.
“No I don’t.”
Luckily, those Cisco routers were around. Not only did routers hook LANs to wide area networks, or WANs, that comprised the Internet, but Cisco routers actually became the backbone of the Internet. New companies like UUNET and America Online would use Cisco routers in the middle of their networks to move packets around, as well as at the edge of their networks to connect to banks of dial-up modems so users could call in and connect.
Marc Andreessen took advantage of these routers. For Cisco, the effect was magic. Browsers blew away the 80-20 rule. They probably flipped it to the 20-80 rule, meaning only 20% of networking was local and the rest had to go through a router to request information and packets from the Internet. Demand for routers exploded. A simple invention, a tiny packet reassembly program named a browser, caused demand to shift rapidly. Without the browser and router growing hand in hand, we’d all still be waiting for the post office to deliver information and orders for products—we might as well have stayed in the industrial age.
Intel and Microsoft, based on Doug Engelbart’s blueprint, put the horizontal into the computer business—thin slices of intellectual property assembled into a final product. The division of labor happened almost immediately. It was cheaper to assemble chips in Malaysia than in Michigan. Cisco’s routers and Netscape’s browsers just rode on top of the computer industry’s platform, and the communications business layered into thin slices similar to the computer business and created another giant waterfall.
Unfortunately for Netscape, Jim Barksdale made every mistake in the book. Microsoft eventually woke up and developed its own Internet Explorer browser, by licensing some of the original code from the University of Illinois. Barksdale’s plan was to acquire his way to success. Many companies began writing add-in software that worked inside of Netscape’s browser. Netscape encouraged this, and there might be five companies doing video plug-ins and six companies doing 3-D code and another four doing database stuff. But Barksdale would immediately turn around and buy one of these companies, in effect freezing out all of the others, who would go work with Microsoft.
I sat in on a presentation by Jim Barksdale at an investors conference soon after the IPO. He set out his master plan. “Look, our model is Cisco. They grow their product offerings via acquisitions. We have a platform that affords us the same model. I am going to do one acquisition a quarter, until I get good at it, and then I’m going to do an acquisition a month.” You could see the investment bankers in the room break into huge smiles. Hello, fees!
Netscape’s stock kept running and running.
I walked into Bob Harris’s office.
“I read today that Jim Clark is now a billionaire,” I told him.
“No he’s not,” Bob shot back.
“Yeah, the stock’s $80 something, and with splits and all, he has 130 million shares.”
“But he’s not a billionaire.”
“Do the math,” I said.
“No, you’re forgetting something important my friend. He’s got to pay taxes.”
“You gotta get over it.”
“No I don’t,” Harris repeated.
But Barksdale’s biggest mistake compares to the Germans invading Russia in World War II and opening a two-front war. IBM was a Netscape ally. IBM had recently completed the acquisition of Lotus, the spreadsheet company. Except Lotus didn’t really make much from spreadsheets anymore; they had reinvented themselves as a groupware company, selling Lotus Notes to corporations to keep track of e-mail and documents and help workers collaborate. Because of this, IBM was very interested in Netscape’s success, almost as the anti-Microsoft. IBM distributed Netscape Navigator on every IBM computer shipped, and resold Netscape server software, a big bonus for Netscape to be able to leverage IBM’s huge distribution prowess. If Microsoft was busy with Netscape, then Microsoft wouldn’t have as many resources to compete with Lotus.
