Tech's Lost Chapter: An Oral History of Boston's Rise and Fall, Part Two

In the modern era of apps, it’s hard to imagine that software once wasn’t considered a business in its own right.

Companies bought or leased room-sized mainframes, and expected the software and services to come along as part of the steep package.

But, under growing antitrust pressures in the late 1960s, the era’s tech giant, IBM, announced that it would “unbundle” software from its hardware business. The decision imbued code with economic value for the first time and, by most estimates, gave birth to the software sector.

It provided an opening for startups to create their own operating systems, databases and applications, and a range of Greater Boston startups seized the chance.

Cullinet became the first pure software company to go public, Software Arts provided the first compelling reason for businesses to buy PCs and Lotus Development grew into one of the biggest tech companies in the world. Along the way, these firms helped invent the way that business software would be developed, packaged, marketed and sold for decades.

But, ultimately, the lessons they offered were learned only too well by upstarts like Oracle and Microsoft on the west coast, and New York’s IBM would soon reclaim some of the territory it ceded.

Here’s how the story unfolded, according to those who were there.

Dan Bricklin, DEC employee from 1973-1976, and co-founder of Software Arts:

You sit in business-school class, and they’re just running numbers. How much do we spend on this? And how much on this? And next year what’s it going to be?

If you’re used to doing that kind of stuff on a blackboard, you say, “well, let’s try programming it.”

My friend Alan, with his programmable calculator, always beat me, even though I’d run down and had this whole DEC PDP-10 downstairs. I thought, “well, maybe if it was only as easy as it is with word processing, if we had word processing with numbers.”

I ended up with a vision of the spreadsheet. There’s a lot of time to daydream when you’ve got 90 kids in your class.

Mitch Kapor, co-founder of Lotus Development Corp:

I bought an Apple II computer in July of ’78, and it changed my life. This was about three years before the IBM PC; it was still strictly a hobbyist market.

But I got a job as a consultant writing programs for other people with Apple IIs, and I helped organize the first Apple II user group in the region, and was just part of the scene.

Everyone knew everyone else. It was a very small community, people were in different roles and relationships with each other.

And those were like the glory days, they were like the early days when nothing had happened yet and everything seemed possible.

Bricklin: I was riding my bike on Martha’s Vineyard and made the decision that when I graduated I wanted to take this idea and try to build some product around it.

There was another student named Dan Fylstra, he was founding a company called Personal Software, here in Allston, Mass. They had a product called Microchess they were selling.

He said, “Whatever you’re doing, you should do it on an Apple or TRS-80.”

So in the fall of 1978, I borrowed over the weekend an Apple II from him, and prototyped in Basic, VisiCalc, the spreadsheet, with the ABC, one-two-three type of thing. We didn’t have a mouse, we used the paddles, so it went left and right, and if you hit the “fire” button it’d go up and down.

But it would recalculate, you could type numbers in, and I had the prototype of the spreadsheet, and that’s when we started.

My friend Bob Frankston and I got together with Dan Fylstra and we made a deal.

Bob and I founded a company called Software Arts on Jan. 2, 1979, which is day one in the counting of time on some spreadsheets.

Bob actually programmed VisiCalc. He started around the winter, and within a couple weeks we had something going. Fylstra eventually took it and showed it to Apple. (Personal Software handled distribution.)

We announced it privately at the West Coast Computer Faire in May of 1979. By then Bob had enough of it going, most of it was working, and we announced it publicly at the National Computer Conference in New York City in June of 1979.

Kapor: I wound up writing a program that was the first graphing and statistics package for the Apple II, called Tiny Troll.

Peter Jennings (co-founder of Personal Software) approached me. He was introduced to me by Bob Frankston. They were interested in acquiring more products to publish, and offered me a contract to rewrite Tiny Troll as a companion to VisiCalc. Eventually, when it finally got to market, it was called VisiPlot.

I worked as a contract program author, under royalty agreement with Personal Software, then I worked for them as a product manager for a couple versions of VisiCalc.

Bricklin: We founded a business and grew it. We started putting VisiCalc on all the computers we could. PCs didn’t catch on that fast, but VisiCalc was selling 1,000 units a month for the first year.

It started getting more and more; others started getting into it, other spreadsheets starting coming in. Apple has acknowledged that VisiCalc was a catalyst, a major catalyst, to the success of Apple in business.

For various reasons, VisiCalc was the only one on the Apple II for a year; the only spreadsheet like it. So that really helped Apple get going in many ways.

But then we had this author-publisher relationship, which was going to be a problem, and the venture capitalists wanted more of this and were paying too much for that. And we ended up in a lawsuit. We were about to sell our company, we were days from selling our company — and we got sued.

Kapor: The real problem was with VisiCalc, and I didn’t want to be collateral damage. They bought me out, and now I had some money; that was in ’81. And that became the basis of Lotus.

Originally, I thought I just wanted to design software. But I concluded that I had to start a business in order to maintain the artistic integrity of the product. I felt that the publishers — and my experience was with VisiCorp — weren’t going to do justice to it.

Now as I look back on it, I kind of backed my way into being an entrepreneur. I had to tell myself a bit of a story because it was out of fashion to start a business if you grew up in the ’60s the way I did. I came of age with “Sgt. Pepper” and the Vietnam War and the Weather Underground.

I was part of the counterculture and all that, so business was very uncool.

Bricklin: I knew about 1-2-3; Mitch showed it to me.

He said, “Hey, look what I’m working on.”

I looked at it and said to myself, “They’re going to sell 60 million in the first year.” They sold 50-some.

Great product. He had been our product manager when he worked at Personal Software, and he knew what makes a good spreadsheet, and he just built the best one he could.

We were trying to be on everything, and because of all our relationships, we couldn’t.

Different people were trying different things. Mitch happened to rifle-shot the IBM PC. So the new generation, 1-2-3, came in, did great, and helped computers get on desks all over the place.

Kapor: I think it’s fair to say, as it is said, it was the killer app that drove adoption of PCs in the business world, globally, in the 1980s. People were buying $5,000 IBM personal computer systems simply to run a $495 spreadsheet.

The spreadsheet was an incredibly brilliant invention by Dan Bricklin, not by me. But what we did do is, we fundamentally improved the spreadsheet in multiple ways simultaneously. So we weren’t a first mover, we were a fast follower.

You could make bigger spreadsheets, and it ran like a bat out of hell, and there was data visualization for the first time. We advertised in Businessweek and the Wall Street Journal, which nobody had really done, and we raised the venture capital to do that. We went out and we trained all the dealers who were selling it; we had a demo disc.

I mean, we just pushed the edge of the envelope in every dimension in a big, huge hurry. So we had a kick-ass product, which was well-marketed.

Reed Sturtevant, designer of Freelance Graphics, a presentation-software package acquired by Lotus in 1986:

What happened with PCs was this kind of democratization of business software. So, if you thought of spreadsheets, word processing, graphics, personal databases, that was a whole new wave and market.

The PC world in Boston started becoming very software-centric.

Ray Ozzie, creator of Lotus Notes, longtime Microsoft executive, and founder of Talko:

In 1982, concurrently with while I was working at Software Arts, I was trying to raise money for a new startup to develop a new product concept (code-named Notes). I failed to get funding. Around that time, my old boss from Data General, Jon Sachs, pinged me and said that he had co-founded a new company with a guy named Mitch Kapor, and suggested that I come over and join them.

Initially I resisted, because I really wanted to do my new startup. But ultimately Jon introduced me to Mitch. After listening to me explain my new product concept, Mitch suggested an idea. He said that if I would come to work at Lotus and develop and ship the next major version of their spreadsheet product, he would personally figure out a way for me to get funded to do the product that I really cared about.

So I joined, and worked with two other great engineers — Barry Spencer and Matt Stern — to develop the product that shipped as Symphony. That was July of 1984. And when it shipped, Mitch was true to his word.

On Dec. 7 of 1984, I founded Iris Associates. Iris was an independent company, funded by Lotus in exchange for a future option to purchase exclusive rights to Notes.

Sturtevant: Lotus was the biggest software company in the world at that time.

We were always, as a startup, worried that we didn’t know what products they were coming out with. If they came out with a competing product before we did, we’d be done.

There were maybe 1,200 employees at Lotus Development. Microsoft at that time in the mid-1980s was maybe 600 employees.

Kapor: The stuff we did became the standard for the PC industry; we defined it. We were the hyper-growth tech startup of the ’80s; we were the ’80s equivalent of Google and Facebook.

Obviously it was smaller, the tech market was smaller, but we had similar numbers in terms of growth. We were the proof that this segment of the market mattered — and mattered to the world generally. We made this product that defined a standard and became ubiquitous. That was the legacy.

Lotus went public in 1983; revenue reached $156 million in 1984, and employment topped 1,300 in 1985. That same year, Lotus acquired VisiCalc, shutting down its flagship product months later; and Microsoft introduced Excel for the Mac.

Bricklin: You could have (imagined) my face when I saw 1-2-3. I knew Mitch was working on something, but I didn’t know what until a few weeks before they announced it.

Years later, when our company basically ran out of money after the lawsuit, we ended up selling what was left of the company to Lotus, because they wanted TK Solver, our other product.

And I was walking through — I think it was Comdex — with Mitch when he saw Excel for the first time, and I saw his face.

It was sad.

Kapor: I don’t remember what I thought when I saw Excel for the first time. It was originally created for the Macintosh, which was not a serious factor in the business market.

What Microsoft did well was to bundle Excel, Word and other productivity applications into Microsoft Office for a single low price, and tie the product closely to Microsoft Windows in the 1990’s when that platform matured.

Lotus was late to market with a good version of Lotus 1-2-3, which allowed Microsoft to gain share as Windows became the dominant platform in the marketplace, displacing DOS.

Sturtevant: If you look at what was happening in the battle of the business software world in PCs, Microsoft made some very, very simple, but very, very strategic decisions.

Microsoft didn’t have the best-selling spreadsheet, didn’t have the best-selling word processor, didn’t have the best-selling presentation-graphics database. None of (their products) were the leader in the category.

They realized you could shift the purchasing decision from “Is it the best software?” to “Is it the best decision for the company?”; bundle the products together, drop the price from $300 or $500 each to $700 for all of them combined; sell to the top of the organization. And they recognized that they could leverage their operating system.

John Landry, chief technology officer at Cullinet and Lotus Development, and vice president of technology strategy at IBM:

Lotus got bamboozled by what’s known as the Microsoft head-fake. Microsoft and IBM had agreed that IBM’s OS2 efforts would be combined with Microsoft to create the operating system of the future, and said, “All you application vendors should write for it.”

Most of the industry bought into that, so everybody starts writing for OS2. Problem is, it winds up coming out and there’s no software written for graphical user interfaces, particularly spreadsheets and word processing.

And shortly thereafter, Microsoft introduces Windows 3 and the entire suite of products: Word, Excel, PowerPoint, etc.

Everybody else is sitting there, holding their you-know-what.

The result is, they got creamed. Excel was so much better. It just kicks ass, and Lotus is sitting there saying, “We’ve got to really rev up a Windows 3 version of 1-2-3 and we’ve got to do it really fast.”

As a result they come out with it, and it’s a freaking disaster.

Bricklin: Lotus just did not get the transition right. And Microsoft spent a fortune working on Excel and getting it on the PC and all this stuff, until eventually they took over with an incredible product.

So Excel was there for the masses. It was designed for the masses, and that’s been the dominant platform for a while.

Landry: Mitch was a transcendental-meditation teacher. And suddenly he finds himself running a public company. As a result, Mitch goes, “Jeez, I need help.” And he brings in McKinsey, the consulting firm. The guy at McKinsey who got the account lead, also with virtually no background in running a company, was Jim Manzi.

One day, not very long after the IPO, Mitch decides that he doesn’t want to run the company anymore. And Jim comes out the other end of the negotiations running it.

Kapor: I gave up the CEO role in 1985, while staying on as chair, and then left that position in 1987. The job had outgrown me and my interest in running a large organization.

Landry: It was awesome from the perspective of being CEO of the hottest company in the PC industry — but then Jim had to figure out what to do as a second product, which was a really tough problem.

They come out with Agenda, and it kind of flops — and another product flops (Metro), and another (Jazz).

Sturtevant: In the early 1990s, if you looked at it as a public company, the revenues were growing slowly. Behind the scenes, I remember a company meeting where they were showing us all a graph and said: “It looks like we’ve been growing slowly. What’s happened is we’ve lost more than $200 million in revenue from our old products in the last couple years and we’ve been working really hard to fill that with our new products.”

Landry, who joined Lotus as CTO in 1990: Why did I want to go there? There was this guy, Ray Ozzie, building Lotus Notes, and that looked really interesting.

But when I got to Lotus, I found out we didn’t actually own Lotus Notes. They let Ray create his own company, Iris. His deal was he would get a percentage of revenue off the top. Lotus would do all the customer support, marketing, sales and have an exclusive on the product, so there was some codependence there.

It turned out that percentage off the top was getting to be a very, very, very big number.

And it gets worse. Lotus did another deal with IBM that was a revenue-split deal, and it didn’t matter who sold the product.

Ozzie: In 1987, Lotus exercised its option to purchase rights to Notes. As a result of that agreement, Iris was to be compensated with a royalty on Notes sales.

Iris was ultimately purchased by Lotus in 1994, essentially converting our royalty stream into Lotus equity.

Landry: We wound up negotiating with Ray and buying him out, so that kind of worked out. But as Notes took off like a rocket worldwide, we were on a good track — and IBM makes a hostile takeover offer on the company.

Jim didn’t want to sell, and that was a problem, because he was CEO. It started a lot of heartache, shall we say, among friends. But eventually logic prevailed, and we got the number up from $60 to $64 a share, and sold the company.

Sturtevant: For the biggest anchor of the software industry in town to be under that much stress really set things back.

Bricklin: Where the power centers are changes all the time. You know, only the paranoid survive, right? And even they don’t always survive.

Kapor went on to co-found the Electronic Frontier Foundation in 1990, and later joined Accel Partners. He remains an active investor and philanthropist through the Kapor Center for Social Impact and Kapor Capital.

Bricklin formed several additional companies, including Trellix Corp., which was acquired, and Software Garden, a consulting and development firm he continues to run. He is also chief technology officer at Alpha Software Corp.

In September 2014, IBM quietly pulled support for the final versions of Lotus 1-2-3, ending a three-decade run with no fanfare.

Lotus Notes turned 25 this last Sunday, an occasion Ray Ozzie marked with a look back via 22-part tweetstorm.

Read Part One of this two-part story here.