Information Processing

Mitch Kapor's Well-Greased Dream Machine

Lotus' founder has $3 million in venture capital to create a new breed of software

Keith H. Hammonds in Cambridge, Mass.
Business Week 

May 30, 1988

Mitchell D. Kapor is back, and he hasn't changed much. The puckish philosopher of the software industry remains slightly paunchy, alternately charming and impatient, and infectiously enthusiastic about his work. But now Kapor is focusing on a new vision. Two years after quitting as chairman of Lotus Development Corp., the No. 2 producer of personal-computer software, Kapor has started a new company with an intriguing, if not fully defined, goal. And he's peppering anyone who will listen with talk of ''fundamentally'' changing the way people use personal computers.

By the time Kapor walked away from the company he founded, its 1-2-3 spreadsheet program had made the analysis and presentation of financial data easier for millions of people. Thanks mainly to 1-2-3, Lotus had become a $283 million operation with 1,300 employees. Kapor had amassed a personal fortune estimated at $100 million.

But for Kapor, Lotus was too big. He knew he lacked the general management skills needed to keep Lotus growing. ''I was trying to find a role for myself,'' he recalls. ''In my heart of hearts, I really wanted to go out and develop products.'' Then, too, he was uncomfortable in an environment too vast for companywide staff meetings--and too insistent on fast returns on investment. He wanted something small and less structured.

So the former meditation instructor is starting again. In a modest Cambridge office overlooking the Charles River, less than a mile from Lotus, Kapor and former Lotus executive Peter B. Miller are assembling a team of top developers for a new venture with the curious title of On Technology Inc.--a name pulled out of thin air.

'GREAT THINGS'

On is informal, echoing the eccentric early days of Lotus, when programmers walked the halls in black satin jackets. But its ambitions are loftier: Through On, Kapor hopes to create software that will help make computers both as simple and as effective a medium of communications as telephones. ''There's a concept here,'' he says, ''which has the potential for doing great things.'' So far, Kapor has bankrolled On with $1 million of his own money. Next month, Kleiner Perkins Caufield & Byers and Cole Gilburne will kick in $3 million of venture capital, even though On will not ship a product until 1990.

Since Kapor wants to keep On relatively small, with no more than 24 employees by yearend, he probably will market his products through joint ventures and licensing agreements rather than an internal sales force. Although he concedes money is important to his investors, it's not critical to him. ''At Lotus, I was interested in building something large and successful. Here, I'm much more interested in accomplishing something important.''

VAGUE

What will On produce? That turns out to be a sticky question, even for Kapor. Usually eloquent and precise, he has struggled in recent speeches to computer customers and engineers to explain On's goals. Instead of specifics, he offers rambling abstractions. It's not that Kapor is being coy. Rather, the concepts are difficult to understand, partly because Kapor himself isn't sure what the applications will turn out to be. ''We've done a lot of brainstorming and scenario formulations about what we might do, but we haven't made the commitments internally,'' he says. ''You might say we don't know what we're doing, and I might agree with you.'' A startup, he argues, need not be bound by rigid strategies.

Kapor can say this much: His On team is developing a software ''platform,'' which will be, in effect, a new layer of software that sits on top of the operating systems, or basic software, for such computers as Apple Computer Inc.'s Macintosh and Sun Microsystems Inc.'s workstations. It will contain sophisticated data-base and communications tools, plus some form of artificial intelligence. These tools, licensed to developers of the applications software that lets computers do useful tasks, will provide programmers with a set of building blocks, much as an interior designer adds walls and fixtures to bare office space.

On's applications programs, or those developed by others using Kapor's ''platform,'' probably will incorporate parts of technologies that already exist. One, for example, may let customers cross-index and annotate electronic copy, so that a word or phrase could trigger further description. Another product may add so-called groupware features that allow real-time editing and discussion of documents over a network of computers. Others may let customers retrieve huge amounts of information from data bases.

Ultimately, Kapor's efforts aim at expanding a computer's role in the workplace by making it easier to use. He foresees the development in the next 15 years of a communications infrastructure that will support a sort of electronic-information marketplace. Such a ''public space,'' as he terms it, might let people have electronic conversations rich with graphic images and symbols.

INTERACTIVE

Kapor's ideas aren't new. Early computer visionaries such as Theodor H. Nelson, Douglas C. Engelbart, and Alan Kay have predicted similar technologies for nearly two decades in the face of much criticism and little commercial acceptance. Five years ago, Brown University's Institute for Research in Information and Scholarship created a project that translates Shakespeare's plays into interactive on-line text. By selecting a certain passage, scholars can get more detail or background. In 1945, Vannevar Bush, of the Massachusetts Institute of Technology, described a futuristic machine that might use a form of artificial intelligence to instantly retrieve data from a huge internal memory.

But Kapor brings commercial expertise to the show. ''He's not an original thinker,'' observes Nelson, who wrote Computer Lib/Dream Machines, a cultish encyclopedia of computer lore. ''But he gets the train into the station.'' Kapor sees On implementing the early theories of others now that the technology is finally becoming available. More powerful personal computers, faster modems for sending computer messages over phone lines, and sophisticated operating systems are encouraging Kapor to pursue his vision: ''There's a window here that's just opening.''

A lot of high-powered talent agrees with him. Kapor's team of 12 includes Miller, a 23-year industry veteran who is On's president, and William A. Woods, an expert in artificial intelligence and On's principal technologist. They joined Kapor, they say, for a shot at the kind of wealth Lotus created for its early employees. But they also share his ideas. ''I've been pushing this sort of thing for four years,'' says R. Stonewall Ballard, an alumnus of Xerox Corp.'s Palo Alto Research Center and a recent On recruit.

On isn't something Kapor had to do: At 37, he already has profited from the entrepreneurial itch as few ever do. With part of the proceeds from his Lotus stock, he established a $10 million foundation to fund research in cognitive science and artificial intelligence--disciplines that analyze and try to replicate human reasoning. He has invested in Go Corp., a computer startup founded last year in San Francisco by former Lotus colleague S. Jerrold Kaplan. Kapor, his wife, psychiatrist Ellen M. Poss, and their two children live on a 22-acre suburban Boston estate.

Kapor was never able to sit still. He founded Lotus after a career considered eclectic even by computer-industry standards (table). After Yale University he worked as a disk jockey for a rock-and-roll station in Hartford. He then moved to Boston, where he got his first programming job, teaching transcendental meditation on the side. Then came a master's degree in psychology and a counseling job in a psychiatric ward.

In 1978, Kapor sold his stereo to buy an Apple II and started programming, supporting himself with free-lance assignments. He succeeded spectacularly: Within three years he had reaped $1.7 million from two statistics and graphics programs. Then came 1-2-3.

IMPROMPTU

It wasn't the first spreadsheet: Daniel S. Bricklin pioneered the concept with VisiCalc. But 1-2-3 was powerful and easy to use, and it worked on International Business Machines Corp.'s new personal computer. Lotus grew quickly in a freewheeling atmosphere characterized by jeans, Hawaiian shirts, and impromptu hallway meetings.

Eventually, it was too impersonal for Kapor. ''He became deeply concerned as the company grew to the size where he didn't know everybody by name,'' recalls Freida Klein, a former Lotus executive. Jim P. Manzi, a streetwise McKinsey & Co. consultant Kapor hired in 1983 as marketing director, quickly took over running Lotus day-to-day. By 1984, Kapor had retreated to strategic planning and software development.

Finally, in July, 1986, Kapor resigned as chairman and CEO. He was paid a retainer to help develop a personal-productivity program for IBM personal computers, called Agenda, which the company will ship next month. But he has sold most of his 14.8% stake. Kapor's strongest remaining link with Lotus may be the legitimacy that his former role has given On. ''Few people have accomplished what he's accomplished, and people respect him for that,'' says Dale Troppito, an early Lotus employee. And Kapor's vision has changed: ''Am I trying to do here what I did at Lotus? Absolutely not.''

Can Kapor pull it off? The On team is nearly finished designing the architecture for its platform. But the applications it will support remain theoretical. There are doubts, even at Lotus, that the project will create anything significant. Kapor takes it in stride, confident he's engaged in something of substance. Besides, he's doing what he enjoys. ''He's having a good time,'' Klein says. ''And he always said that that's what was important to him.''

Photograph: KAPOR: ''YOU MIGHT SAY WE DON'T KNOW WHAT WE'RE DOING, AND I MIGHT AGREE WITH YOU'' BRIAN SMITH

MITCH KAPOR'S JOURNEY

1967-71At Yale, he studies psychology, linguistics, and computer science and manages the campus radio station. He also develops an interest in Eastern mysticism

1973 Moves to Boston as an entry-level programmer. Teaches transcendental meditation

1978 Gets master's degree in psychology and works as psychiatric counselor. Buys an Apple II computer

1979 While a free-lance programmer, he writes Tiny Troll, a statistics and graphics program, for a business-school friend

1981 Revises Tiny Troll for VisiCorp, earns $500,000 in royalties, then sells rights to it for $1.2 million

1982 With programmer Jonathan Sachs, he writes the first version of the 1-2-3 spreadsheet program. Obtains financing from venture capitalist Benjamin Rosen and forms Lotus

1983 Hires Jim P. Manzi, McKinsey consultant, as marketing director. Lotus has first-year sales of $53 million

1984 Lotus names Kapor chairman and Manzi president.

Company grows to $157 million in sales, 700 employees 1986Resigns as chairman and chief executive. Negotiates a one-year deal to finish Lotus' Agenda, a personal productivity program 1987With Peter B. Miller, he founds On Technology Inc. DATA: BW

Copyright 1988 McGraw-Hill, Inc.