Dan Bricklin and VisiCalc: The Killer App

Zusammenfassung

In 1979, a Harvard Business School student named Dan Bricklin shipped a program that transformed the personal computer from a hobbyist toy into a business tool. VisiCalc — the “Visible Calculator” — was the world’s first electronic spreadsheet, and it did something no program had done before: it gave ordinary people a reason to buy a computer. Accountants, financial analysts, and small business owners bought Apple IIs by the hundreds of thousands specifically to run VisiCalc. Steve Jobs later credited it as the program that legitimized the personal computer industry. The bitter irony of Bricklin’s story is that he invented the concept, built the product, transformed an industry — and ended up with very little to show for it financially, outmaneuvered by imitators who arrived after the market he had created was proven.

The Tedium That Launched an Industry

The idea came from watching a professor.

In the spring of 1978, Dan Bricklin was a student at Harvard Business School. In a case study class, he watched his professor work through a financial model on a blackboard: a grid of numbers, rows and columns, with relationships between cells. When one number changed — an assumption about interest rates, or revenue growth — every dependent cell had to be recalculated by hand, then re-entered on the board. The professor erased, recalculated, rewrote. The process was slow, error-prone, and fundamentally stupid. The numbers in every cell were not independent; they were related by formulas. A machine could track those relationships and recalculate automatically.

Bricklin had an MIT computer science background before Harvard — he had worked at DEC and published research on text editing. He understood immediately that what he was imagining was a programming problem with a known shape: a two-dimensional array, with cells that could contain either values or formulas referencing other cells, with automatic recalculation propagating changes through the dependency graph. He prototyped a version in BASIC on an Apple II that fall.

What he needed was someone who could make it fast enough to be useful. He found Bob Frankston, a MIT friend who was one of the best programmers either of them knew. Frankston rewrote the prototype in 6502 assembly language — the native machine language of the Apple II’s processor — with an obsessive focus on performance. The finished program needed to fit in 20 kilobytes of RAM (later versions grew to 28KB) and respond to keystrokes instantly. Frankston worked on it through the winter of 1978–1979, often in the early hours of the morning when he had the office to himself.

Software Arts and the Apple II

Bricklin and Frankston founded Software Arts in January 1979 to publish the program. They signed a distribution agreement with Dan Fylstra’s Personal Software company, which handled marketing and sales. The arrangement would later prove costly: Personal Software (which renamed itself VisiCorp) controlled distribution and the VisiCalc brand, while Software Arts owned the software but had limited leverage over how it was sold.

VisiCalc — the name combined “visible” and “calculator” — shipped for the Apple II in October 1979 at a retail price of $99.95. The program displayed a two-dimensional grid of rows and columns. Each cell could contain a number, a label, or a formula. Formulas could reference other cells: typing +A1*B2 in cell C1 would display the product of whatever values were in A1 and B2, updating instantly whenever either changed. The spreadsheet recalculated from top-left to bottom-right, and the screen updated in real time.

The response was immediate and astonishing. VisiCalc sold 1,000 copies in its first month. Within a year, it was selling 12,000 copies per month. By 1983, Software Arts had sold more than 700,000 copies across multiple platforms. The Apple II version was the biggest seller, and accountants and financial professionals who had never owned a computer began buying Apple IIs — at roughly $1,500 each — specifically to run VisiCalc.

“The $99 Accounting Package That Sells a $1,500 Computer”

The phenomenon was noticed immediately by the business press. Inc. magazine ran an article in 1980 describing small business owners buying Apple IIs they didn’t fully understand because VisiCalc justified the cost. A controller at a Boston firm told the magazine he had done in three hours with VisiCalc what previously required three days of manual calculation. Steve Jobs, who understood the dynamic precisely, later said: “VisiCalc propelled the Apple II to the success it achieved more than any other single event.” For the first time in computing history, software drove hardware adoption rather than the reverse. The “killer app” — a single program so useful it justifies purchasing an entire platform — was a new phenomenon, and VisiCalc defined the concept.

The Platform Problem

VisiCalc’s success attracted imitators, and the legal landscape around software at the time offered Bricklin and Frankston almost no protection.

Software patents, as they would later be understood, did not yet exist in a form that could protect a user interface concept like the spreadsheet. Copyright protected the specific code Frankston had written — but not the idea of a grid-based recalculating spreadsheet, which was an abstract concept. Any competitor could write their own version of the same idea from scratch without infringing.

Bricklin consulted patent attorneys and was told that software methods were not patentable under prevailing law. (The Supreme Court’s Diamond v. Diehr decision in 1981 began to shift this, but too late and too narrowly to help Software Arts.) The concept that Bricklin had originated — the electronic spreadsheet — was freely available for anyone to implement.

The distribution agreement with VisiCorp compounded the problem. By controlling VisiCalc’s distribution channel and brand, VisiCorp had leverage that Software Arts lacked. When VisiCorp began developing its own spreadsheet product (VisiCalc Advanced) and later attempted to acquire Software Arts outright, the negotiations were adversarial. Bricklin and Frankston found themselves in legal disputes with the company that was supposed to be selling their product.

The Lotus Coup: 1983

The competitor who ultimately displaced VisiCalc was not a minor imitator. It was Lotus 1-2-3, released in January 1983 by Mitch Kapor and Jonathan Sachs for the IBM PC.

Lotus 1-2-3 was faster than VisiCalc (written in tight assembly language specifically for the 8088 processor), integrated charting and simple database functions into the spreadsheet, and had a consistent command structure more polished than VisiCalc’s. Crucially, it was designed from the ground up for the IBM PC — the machine that was rapidly displacing the Apple II in business settings. IBM PC sales in 1983 dwarfed Apple II sales among the corporate customers who were VisiCalc’s core market.

Lotus 1-2-3 reached $50 million in sales in its first year. Within eighteen months of its release, it had displaced VisiCalc in virtually every business computing context. VisiCalc’s sales collapsed. Software Arts attempted to release a competing product — TaxCalc, then a more capable spreadsheet — but could not match Lotus’s momentum or resources.

In 1985, VisiCorp and Software Arts settled their legal disputes and Software Arts was effectively dissolved. Lotus acquired the VisiCalc name and intellectual property. The company that had created the spreadsheet category was gone six years after its founding.

For the continuing spreadsheet story — how Lotus itself was later displaced by Microsoft Excel — see The Spreadsheet Revolution.

What Bricklin Got (and Didn’t Get)

Dan Bricklin did not become wealthy from VisiCalc. The $99.95 retail price generated substantial revenue, but Software Arts’ share — after VisiCorp’s distribution cut — was limited, and the company spent much of it on engineering and the legal disputes with VisiCorp. After Software Arts dissolved, Bricklin held no equity in the surviving entity and received no ongoing royalties from Lotus or its successors.

He had invented a concept worth billions of dollars in aggregate and collected very little of it. The bitterness this might have generated was, by most accounts, largely absent from Bricklin’s public presentation. He wrote and spoke extensively about VisiCalc’s history with evident pride in what the product had accomplished. His explanation for the experience was characteristically analytical: he had made business and legal decisions that seemed reasonable in context, in a legal environment that did not protect the kind of innovation he had made.

What he did collect was substantial recognition. The industry understood what VisiCalc had been. The Computer History Museum inducted him into its Fellow program. The ACM Software System Award for 1999 went to VisiCalc, shared by Bricklin and Frankston. Business schools used the VisiCalc story in case studies on first-mover disadvantage and the limits of intellectual property protection.

After VisiCalc

Bricklin did not retire from software after Software Arts. He founded Trellix Corporation in 1995, building web-publishing software, which was acquired by Interleaf. He then joined Alpha Software, a database software company, where he worked for many years as a senior fellow.

He became an articulate advocate for open source software and the “gift economy” of the early internet, writing and speaking about how the failure of traditional intellectual property protection had shaped his thinking about software creation. He released several of his later projects — including SocialCalc, a browser-based spreadsheet he co-developed with the Socialtext team — as open source. SocialCalc became the foundation of the spreadsheet component in various web applications, including the OLPC (One Laptop Per Child) project.

The VisiCalc source code itself was released to the public domain by Bricklin in 2012 — a generous act toward historians and educators that produced exactly no financial return.

The Inventor Problem

Bricklin’s story illuminates a structural problem in the technology industry’s relationship with inventors: the gap between creating a category and capturing its value.

The economists call it the appropriability problem: innovators may create enormous social value without being able to appropriate more than a small fraction of it for themselves. VisiCalc created the spreadsheet category, made the personal computer relevant to business, and drove hardware sales worth hundreds of millions of dollars. The appropriable portion of that value — the revenue Software Arts could capture — was limited by the absence of patent protection, the distribution agreement with VisiCorp, and the speed with which better-capitalized competitors could enter a market the product had proven.

Bricklin saw this clearly. “I created a category, and someone else took it,” he told interviewers. “That’s how it works sometimes.” The equanimity was real, but the lesson was not lost on the software industry. The era of software patents that followed — whatever its other problems — was partly a response to the demonstration, provided by VisiCalc, that the most consequential software innovations could be created and immediately expropriated.