Cambridge, Mass. — It's a page right out of computer folklore. It's February 1983. Two guys, aged 24 and 34, head off to a cabin in Vermont. They go armed with 25 pads of legal paper, lots of peanut butter, firewood. A few weeks later they come out with an idea that they believe will revolutionize the business world. Now that idea is etched on a floppy disk. It's called Javelin, a business-analysis software package for microcomputers.
The final chapter -- indeed, the body of the story -- has not been written yet. There will certainly be drama: The new company is being born at a stormy time in the computer industry, when many small companies are failing or being bought, and even the biggest ones are stumbling a bit.
But the founders of Javelin have had some good signs thus far. They got financing at a time that venture capitalists had cut back the number of high-tech start-ups they funded by 45 percent. They persuaded 25 of the biggest companies in the United States to test and help revise their product. Some of those, such as Merrill Lynch and BankAmerica, swear by Javelin.
Javelin was also given a lot of press well before it unveiled its package this week. Enough press, in fact, to put the No. 1 microcomputer software company, Lotus Development Corporation, somewhat on edge. Of course, that has its disadvantages. Lotus (which, like Javelin, is based in Cambridge, Mass.) has been alerted to the Javelin threat, and Lotus has the manpower, reputation, market dominance, and dollars to wage a pretty tough war.
``If Lotus does nothing to respond within a year, it will be in sorry shape,'' says Curt Monash, first vice-president at PaineWebber. ``But the chances of that are between slim and none. . . . I'm quite sure Lotus has enhancements to spreadsheets on the shelf that would transform 1-2-3.''
Lotus 1-2-3 has become the industry standard in spreadsheets. Spreadsheets are basically a computerized pencil and paper -- but much faster and, if the right numbers are plugged in, more accurate. They are often used to make forecasts and plot business strategy.
For more than two years, 1-2-3 has topped the best-seller list of software products, having gone to an estimated 800,000 to 1 million users. But even 1-2-3's sales have slowed as shipments of microcomputers tailed off this year. Many analysts say that after micro sales pick up, the nature of the software market will have changed.
``You won't ever see a product with the stunning success of 1-2-3,'' which appealed to companies across the board, says William Zachmann, corporate vice-president of research at International Data Corporation. ``It's a fragmented market now,'' he says, and it's smarter to specialize in software for specific types of companies, like financial firms, car manufacturers, etc.
Javelin is not only going after the across-the-board sales, but must also get customers to switch from Lotus, which has a lock on 65 percent of the spreadsheet market. Javelin is unlikely to displace Lotus, says Jay Owens at Arthur Andersen & Co., the accounting firm, which tested an early version of Javelin. It has a specialized application, financial modeling, while Lotus offers more-general features (like compiling mailing lists).
The market is potentially very big, Mr. Owens says, but ``the majority of companies, both large and small, don't do sophisticated financial modeling.''
How will Javelin lure sophisticated customers from Lotus? Certainly not by price: Javelin sells for $695, while 1-2-3 retails at $495. The main selling point, says Stanley Kugell, president and cofounder, is that Javelin is easy to use and ``it understands business principles.''
Spreadsheets have a way of entangling the user in the process of the analysis, putting in the numbers and writing the assumptions about the program in computerese. Usually the ones doing this work are inexperienced analysts, says Robert Firmin, Javelin's chairman and cofounder, and managers often do not go back to check the analysts' assumptions.
``The result is you have a great gap, and the analyses don't reflect the business experience and judgment of middle and senior management,'' Mr. Firmin says.
Javelin allows one to write in the assumptions in natural language, leaving them in the program, ready to be checked. Javelin also lets you look at and alter the analysis of, say, projected earnings, several ways: through bar charts, pie charts, line graphs, worksheets, tables.
Javelin the company is not pinning all its hopes on Javelin the business-analysis package, Firmin says. Javelin can easily be tailored to vertical markets, he says, and those types of products should be rolling out in the first half of 1986.
Mr. Kugell is guiding Javelin's research in other areas of software. At 26, he has some experience with computers: He joined MIT's Artificial Intelligence (AI) Lab at age 12, went to Stanford's AI Lab at 14, and two years later skirted child-labor laws to join the research staff at Xerox's Palo Alto Research Center.
Javelin is the third company he has started. The company has in-house experts working on software for minicomputers and mainframes. The fruits of that research will probably not be seen in the next year, Firmin says.