Overdrive, page 21
By the fall of 1994, when thousands of computer users began downloading Netscape’s new browser, the number of people using the Internet was exploding by about 10 percent a month. Many of these new users were interested in just one area of the Internet—the World Wide Web. From some 50 commercial sites in January 1993, by October the Web had about 10,000. The number was doubling every few months, and the volume of traffic was doubling every four months. More than 600,000 copies of NCSA Mosaic had been downloaded for free, and upward of 80,000 copies a month were still being grabbed off the Internet.
The Web represented a hotbed of innovation and a golden opportunity for the big three commercial on-line services, which combined had about 6 million subscribers. CompuServe, Prodigy, and America Online were all scrambling to find partners with browser technology that would provide their customers with quick and easy access to the Internet. Although Microsoft would not officially announce that it was developing an on-line service until the Comdex trade show in November, it had been negotiating with two companies, Book- link Technologies and Spyglass, about licensing their browsers. Spyglass was marketing the Mosaic browser that had been developed by Marc Andreessen and his team at the National Center for Supercomputing Applications (NCSA). Spyglass had licensed Mosaic to about 20 companies, which in turn
planned to peddle about 12 million copies. Booklink, on the other hand, was a tiny company with about a dozen employees; nevertheless, its browser technology was also attracting interest.
Gates and his technical assistant Steven Sinofsky had discovered Booklink’s browser when they visited the company’s booth at the spring Comdex in Atlanta in late May, about a month and a half after Microsoft’s Internet retreat on April 5 at the Shumway Mansion. Gates spent 30 minutes talking about the browser with David Wetherell, who had founded Booklink in February. “We had a good talk at Comdex, and after that we started talking with Microsoft about a lot more things,” said Wetherell.
One product idea the two companies talked about was a so-called Internet assistant that would enable users of Microsoft Word to create pages for the World Wide Web, an idea also raised during the April retreat.
Following Comdex, Sinofsky showed the Booklink browser to Brad Silverberg, vice president of Microsoft’s operating systems, including Windows 95. Russell Siegelman, who was directing the development of the Microsoft Network, also got involved.
What really jump-started Microsoft’s decision to build a browser was a memo that senior programmer Ben Slivka sent to his superiors Silverberg and John Ludwig on August 15, strongly urging that Microsoft assemble a browser development team. But browser technology was not yet a priority at Microsoft. To date, only five programmers had been assigned to help Slivka with the project, and it would be early 1995 before Microsoft began to throw more resources into its own effort to build a browser.
The 33-year-old Slivka was a veteran at Microsoft, having worked on DOS 6.0 and Windows 95. By mid-1994, his responsibilities for Windows 95 were winding down. Much of the code had been written, and debugging was in progress.
Slivka’s post-Windows mission was to help figure out a strategy for merging Windows with Windows NT and to come up with recommendations for new features that could be incorporated into the next version of Windows. In examining technology trends, Slivka had begun playing around on the Internet. It soon became apparent to him that the Internet could eventually take the place of Windows as the dominant personal computer platform. The key to tapping the potential of the Internet was a browser, and Microsoft wasn’t even working on one. “I wanted to build one, just because it was a fun thing to do more than anything else,” said Slivka. “The way it works at Microsoft is if you have an idea and can convince people it’s a good thing to do, you can get to do it.”
With so few people on the project, it was impossible for Microsoft to build a browser from scratch and have it ready for Windows 95. Booklink’s browser technology suddenly became much more attractive. In September, about a month after Slivka wrote his memo, Microsoft began serious talks with Booklink’s Wetherell. Siegelman, a crack negotiator, was Microsoft’s point man in the talks.
“Our browser was the best in the world at the time,” said Wetherell. “Microsoft said they wanted to license it for Windows 95 and for the Microsoft Network. We got very close to making a deal, but they did not sign it.”
Early in the negotiations, Wetherell learned what other companies that had done business with Microsoft already knew: Gates does not do royalties. Only rarely has Gates agreed to a licensing contract stipulating that Microsoft pay royalties for a product—although Gates has no problem with contracts stating that the other party pay royalties to Microsoft. Siegelman offered Wetherell a one-time, flat fee of $2 million for the source code to Booklink’s browser. Wetherell thought—rightly, it turned out—that the figure was much too low. CompuServe would later pay $100 million for a competing browser. “Microsoft said they would not pay more because there was no money to be made in browsers,” Wetherell said. “They thought of it as a commodity. Well, if that’s true, isn’t e-mail a commodity? So shouldn’t Exchange be free? Isn’t the operating system a commodity? Obviously, there was a lot of money to be made in browsers. And Netscape proved it. At the time, I thought Microsoft was making a big mistake in their thinking. There was a lot of money to be made, as Microsoft found out later, by giving away things for free.”
Although Microsoft would not budge on the $2 million figure, Siegelman did throw in an incentive with potentially a great deal of value for Booklink. Microsoft would give Book- link the electronic rights to a portion of the Microsoft Bookstore, whose content it could market on the Microsoft Network. This sounded attractive to Wetherell, though he was worried about Blackbird, the project that was under way to develop proprietary tools for putting content on Microsoft’s Network. Microsoft was hoping that Blackbird would become the industry standard. But Wetherell wondered: What if Blackbird doesn’t take off? To which Siegelman replied that that was a risk Booklink would just have to take. “They weren’t giving us money to convert their content or to do the conversion ourselves and then market it at the risk that Blackbird would not take off as a standard,” Wetherell said. “And as it turned out, Blackbird didn’t take off.”
By early November, Microsoft and Booklink were just days away from signing a licensing deal when another suitor came calling for Booklink’s browser—American Online (AOL), which at the time already had more than 1.25 million subscribers and was expanding its on-line content daily. Although it still trailed Prodigy and CompuServe, AOL was the fastest growing of the three services and was clearly aiming to be number one. It had established strategic alliances with dozens of companies, including Time Warner, ABC, NBC, Knight-Ridder, IBM, and Apple. More important, AOL had devised an Internet strategy and was aggressively making the acquisitions necessary to implement that plan.
“AOL came in and made us a very good offer,” Wetherell said. “We did not go back and forth with Microsoft. It was just simply that AOL made us a terrific offer, and that offer materialized very very quickly. And we accepted.” Under the terms of the contract, Booklink’s parent company CMG received 710,000 shares of AOL’s common stock, valued at the time at $30 million, for all outstanding Booklink stock. It turned out to be a great deal, financially, for CMG. During the negotiations with both AOL and Microsoft, Wetherell had been convinced that AOL’s stock was undervalued, which was one reason he spurned the offer from Microsoft. Even though AOL’s stock had been in a free fall caused by investors who were worried about the competitive impact of Microsoft’s on-line service, Wetherell figured that it would go back up. He was right. That $30 million grew to $73 million within 10 months as the price of AOL’s stock skyrocketed. Not a bad profit on a $1.5 million investment made by Booklink to develop its browser.
America Online had promised that it would provide Internet access to its subscribers by the first quarter of 1995, and the swiftness of AOL’s move to snatch away Booklink’s browser caught Microsoft by surprise. “That woke us up. We had to be a lot more aggressive, a lot more lively,” Brad Silverberg would later say.
But, according to Wetherell, Microsoft could have sealed the deal long before AOL came calling. “Once AOL entered the picture, Microsoft was scrambling to try and get the contracts ready. But they sat on the contracts for a long time. Had they moved like they usually move, which is very swiftly—they are the most agile company I’ve ever done business with—the deal would probably have been done. But AOL was incredibly agile, too.”
Wetherell suspected that AOL was less interested in Booklink’s browser than it was in keeping the technology away from Microsoft. AOL’s president, Steve Case, was increasingly worried that the Microsoft Network, bundled with Windows 95, would present a formidable challenge to his company. About the same time that AOL won Booklink, it also completed an acquisition of NaviSoft, a developer of software products to help content companies. By using NaviSoft’s tools, content companies would be able to launch Internet services such as catalogs, classified ads, or journals without writing a single line of code. America Online also spent another $35 million for Advanced Network & Services, developer and operator of a high-performance data network known as ANSnet for business, research, education, and government organizations. These three acquisitions formed the nucleus of AOL’s Internet Services Company, which was put under the direction of David Cole, a venture capitalist and former chief executive officer of NaviSoft and president of Ziff Communications and Ashton-Tate.
Before buying Booklink, Case seriously considered licensing Netscape’s browser, but Netscape would not commit to the kind of deal Case wanted. “They were eager to license to us, but not on the terms that we thought would be appropriate, or to make the level of commitment in terms of customizing their technology to meet our needs that we felt was essential, so no deal was ever struck,” said Case, who had talked with venture capitalists Kleiner Perkins Caufield & Byers about investing in Jim Clark’s company shortly before it was founded in April. Nothing was ever finalized, and Kleiner Perkins found other investors.
America Online wanted to embed a browser within its application, rather than have the browser separate. Netscape’s technology, however, was not as modular as Booklink’s and therefore would have been more difficult to embed. After a careful technical review, Case decided it made more sense to acquire Booklink than to license Netscape’s browser. “We thought that launching a browser and launching a separate application from within AOL, would nullify the interactive experience we were trying to create,” said Case. “And obviously Netscape had a lot of things on their plate, including entering the corporate market. We didn’t feel that they were going to be able to make us the priority that we felt would have been necessary.”
Case said that AOL had multiple objectives in going after Booklink, and keeping the technology away from Microsoft was certainly one. “We recognized the growing centrality of the Web and examined a variety of different Web technologies,” he said. “We talked to a dozen different companies, and Booklink stood head and shoulders above the pack, in terms of both their product and their technical team. Initially, we were having discussions with them about licensing the technology, but the more we walked down the path, the more we felt it would be better to own it. And, absolutely, part of that motivation was a sense that Microsoft was likely to do a deal with them. Microsoft was considering an acquisition, and we had a lot of concerns that Microsoft could accelerate their entry into our market by piggybacking on Booklink’s technology. So the acquisition really gave us some Web technology overnight, gave us some additional insights into the evolving Web market, and preempted Microsoft from acquiring it.”
Just days after the deal, at the November Comdex show, AOL’s newly acquired Booklink browser received a public relations boost when it was named Rookie of the Year by Byte magazine, an award given each year at the trade show to recognize the hottest new company in the computer industry.
Although Microsoft lost out on the browser, it did sign a license agreement for Booklink’s Internet technology to be used with Microsoft Word. Booklink developed for Microsoft a product called Internet Assistant for Word, which automatically generates hypertext markup language, or HTML, the standard Internet file format, freeing Word users from having to learn the HTML formal and manually input the HTML pages needed to create Internet documents.
On November 18, Prodigy became the first of the top three commercial on-line services to provide access to the World Wide Web. Ten days later, CompuServe linked up with a Seattle company, Spry, which was selling a version of the popular Mosaic browser. CompuServe, owned by H&R Block, paid a staggering $100 million in cash and stock for Spry. The Internet market had become red-hot. In late November, MCI Communications Corporation announced that it was teaming with Netscape to offer Internet access and software by January 1995. It was the first major success for Netscape. Microsoft, meanwhile, was in serious talks with Spyglass about its browser. Ironically, months earlier, a Spyglass executive had been rebuffed by Microsoft when he called to ask if Microsoft might be interested in licensing Mosaic. Microsoft had told him that it would build its own browser. Now it was Microsoft that wanted to make a deal with Spyglass. And it turned out to be the most critical deal Microsoft would make as it rushed to catch up with Netscape and the rest of the industry that was now running on Internet time.
The phone call that Michael Tyrrell had been hoping for came early in the afternoon of Friday, November 11, 1994. Tyrrell, executive vice president of business development for Spyglass, was in his office in Cambridge, Massachusetts. For the past several months, he and Microsoft had been going back and forth on whether Microsoft would license the Spyglass browser for Windows 95. But after American Online reached an agreement with Booklink for its browser, the talks had begun to heat up. Suddenly, Microsoft was showing much more interest. On the phone that morning was Thomas Reardon, program manager for Microsoft’s browser development group and the company’s principal representative in the talks with Spyglass. Reardon worked with Ben Slivka, who had written the August memo urging Microsoft to have a browser ready to go with Windows 95.
“We are ready to talk,” Reardon said. “We’d like you and Tim [Krauskopf, founder and vice president of Spyglass] to be out here first thing Monday morning. We want to get this deal done by next week.” Tyrrell responded that he did not want to fly all the way out to Seattle for a meeting unless Microsoft was serious. “We are very serious,” Reardon told him. “We want to move, and move fast. We want to get this done.”
After Reardon hung up, Tyrrell phoned Spyglass headquarters in Naperville, Illinois. He wanted to give the good news to Krauskopf. “Get your bags packed,” Tyrrell told him, “we have to be at Microsoft first thing Monday morning. They are ready to make a deal!”
Both Tyrrell and Krauskopf realized just how important an agreement with Microsoft would be for the future of Spyglass. A few months earlier, in August, Spyglass had signed a contract with the National Center for Supercomputing Applications that gave Spyglass the master license to sell the Mosaic browser that had been developed at the center, meaning that any company that wanted to license the browser now had to come to Spyglass. Microsoft had wanted to deal only with the center, not with Spyglass, but now, if it wanted Mosaic, it would have to negotiate with Tyrrell and Krauskopf.
“This was such an important contract to Spyglass,” said Tyrrell. “Netscape had already released its browser, and it had more features and they were getting more notoriety than we were, by leaps and bounds. This deal with Microsoft was everything to Spyglass. It was everything.”
For Krauskopf, it was the realization of an entrepreneurial dream that had started when he was a student at the National Center for Supercomputing Applications, one of five big centers around the country that had been funded by the government to help research scientists using supercomputers, in much the same way that Tim Berners-Lee had been given a charter to help high-energy physicists at CERN, the research lab in Geneva. Berners-Lee had invented the World Wide Web as a way for scientists to exchange information. In the mid- 1980s, as part of the center’s brain trust under director Larry Smarr, Krauskopf and another student developed NCSA Telnet, which became the standard communication protocol for logging in to a supercomputer with a PC or Macintosh. NCSA Telnet was the first experiment in “freeware” that brought the Internet to every PC and Mac. Today, it is used by every university in the country. Mosaic was the next free software developed by the university, and it brought the Internet to the masses.
Telnet made it possible for scientists to transfer millions of numbers from supercomputers to personal computers for the study of astronomy, hydrodynamics, and other fields. But they could not do anything with all those numbers without the software to display them as images. Krauskopf wrote a series of scientific visualization programs, and in 1990, with some prompting from friends and a loan from Apple Computer, he founded Spyglass to commercialize that visualization software. The company was initially headquartered in Champaign, Illinois, home of the supercomputing center at the University of Illinois. “At the time, scientific visualization was very hot and we were going to grow that product line to become profitable and establish the company,” said Krauskopf. “We wanted to go out and capture a certain portion of the market for people who needed to do imaging and analysis of scientific numbers, and we achieved that. At that point, we had not as a company embraced the Internet at all. My ties to the Internet were very solid, but that’s not what the company was doing. The Internet was very useful for universities, but it had not yet branched out into commercial applications. Most scientists could not see how to do their jobs without it. But it was still very much for research only.”
