This month the spreadsheet turns 25.4. Or 23.9. Or 24.7. That’s the beauty of the spreadsheet: change your assumptions and the numbers that derive from them change as well. So whether you’re dating the spreadsheet’s creation from its initial conception, its first commercial release, or the incorporation of the company that brought it to market, the ramifications of a new approach to a problem can be calculated instantly. To ask what-if is to know “What if…?”
Actually, that’s just one of its many beauties. In the quarter century since Dan Bricklin first conceived of the idea of the electronic spreadsheet, the software program has proved to be astoundingly flexible and nearly ubiquitous, used by people in every conceivable walk of life to not only “run the numbers” but also manage data in scores of ways, many of them unforeseen even by its creators. After all, in that spring of 1978 when Bricklin first imagined the product that would come to be known as VisiCalc, all he wanted to do was ace business-school case studies. And then came the Pepsi Challenge.
As a Harvard MBA student, Bricklin was assigned a case study built around Pepsi’s marketing campaigns. At a time when students relied on hand calculators and ratios to crunch the numbers, he was able to offer a series of very specific projections built around many variables. Asked by the professor how he managed to present so much detail, Bricklin danced around the issue, in part to keep the product a secret (he already knew where his postgraduate life would take him) and in part to avoid having to explain that the early program code didn’t always compute. But he and co-creator Bob Frankston soon worked out the bugs, formed Software Arts Inc., and saw their product become one of the two undisputed “killer apps” (the other being word processing) that gave people a reason to own a personal computer. Now chief technology officer at Web-hosting company Interland Inc., Bricklin sat down with CFO IT editor Scott Leibs for a look back, and ahead.
How is it that a program created in the earliest days of PCs, inspired at least in part by the immediate need to tackle a class assignment, has become a staple not only of the business world but also of life in general?
A lot of it depends on the basic structure of the spreadsheet. When you work with numbers, a grid of some sort proves to be extremely useful. The electronic spreadsheet took basic calculating, simple numerical computing, and other functions and married them with a layout for expressing and presenting results. It’s a very powerful combination of restrictions and freedom. It took the calculating aspect and the input/output and combined them in a very natural way. It’s a general-purpose tool, and in fact, as VisiCalc gave way to Lotus 1-2-3 and then to Microsoft Excel, those products further emphasized the use of columns and rows as a simple database. That was a major development; it made the spreadsheet a word processor or desktop publisher of financial information.
Even so, are you surprised that, given the incredible pace of technological innovation, the spreadsheet lives on?
General-purpose tools are really hard to knock off. Word processors can do everything, right? Anything you want to write, more or less, you can write with a word processor. The same is true of spreadsheets, but it’s not true of many other programs.
With computers, you like to learn one tool and learn it well and customize it to your needs. Talk to the people at Google: the broad nature of the search requests they get is amazing. Britney Spears may be the most searched-for name, but that represents just a tiny fraction of what people search for. Everyone has individual needs, and general-purpose tools that can be customized play a huge role.
The fact that you can modify it is extremely useful. How many CFOs inherited the spreadsheet of their predecessor and then modified it?
Interesting that you’d cite word processing. I imagine not many people know that you actually helped develop word-processing programs before you began to develop VisiCalc.
It’s not surprising that given my background I’d come up with something like VisiCalc. Others probably would have, too, and others actually did come up with the general idea; I just made sure mine came to market! My technology background [Bricklin attended the Massachusetts Institute of Technology as an undergraduate] combined with a knowledge of interactive computer languages, word processing, and what I learned as an MBA student all played a part. I was not just looking for financial forecasting; I was also looking for production planning and for how to do marketing-oriented calculations, so again it comes back to that idea of a general-purpose tool.
You left out the part about Watergate.
Right. Early on I worked on software for typesetting and wire-service systems. Once I was flying back from Kansas City after helping them with a new system, and I had to turn around midroute and go back because the transcript of the Watergate tapes had just come out and we had to be sure my program could handle what the wire services were sending out.
But fortunately the mood in Cambridge was decidedly different from that in Washington.
Yes, there was this opportunity of coming from the computer world and knowing this other, working world: I and others felt that we could bring something from one to the other. There was sort of an imperative there that you can improve the world by taking from something that’s not so obvious and moving it to [the mainstream]. The view was, if you don’t do it, who will? Nowadays everybody will do it, but in those days, that wasn’t the case. Computers were seen as strange and weird things, meant for people in lab coats.
And yet VisiCalc was originally seen as a viable product for the consumer market.
Yes, Dan Fylstra, who headed up Personal Software, the company that published [distributed] VisiCalc, was looking for a program that could balance checkbooks. Of course, it turns out in hindsight that one doesn’t want a checkbook program, one wants a “how do you pay the bills easy” program, and certainly [personal-finance software maker] Intuit understands that. But the fact that the spreadsheet could be used as a checkbook got some early attention, although it soon became apparent that it could do lots of other things, and that it was suited to professional accountants and businesspeople, not just home users.
In fact, some aspects of its basic design presume a business use?
Yes. For example, if you ask why the early versions had 63 columns, it’s because if you’re doing material resource planning of any sort—any kind of production planning—it’s done on a weekly basis. You like to do things on a weekly basis for a year or beyond, so despite limited computer memory, we decided on 63. And it had the look-up function because Bob [Frankston] was trying to do taxes with it, and for taxes you need to be able to do the look-up types of stuff. We used decimal arithmetic in large part so it would round in the same way handheld calculators did, and we put in the ability to draw a horizontal line because you need them to create balance sheets and similar reports.
Looking back, is there anything that you wish you had done differently?
Well, if we had had a better business relationship with our publisher, we might still be dominant now, but a lot of people can say that. Also, at the time you really had to write software for specific computers; you couldn’t write a fast program that would run on everything, and we wrote Version 2.0 for Apple. Lotus obviously went with IBM, and we know how that worked out.
From its earliest days, the spreadsheet was immediately seen as useful for things other than financial calculations. Who were some of your first nonfinancial customers?
As I recall, among the first hundred users were people in the medical field, who used it to calculate factors pertaining to anesthesiology and open-heart surgery. When you find out something like that, you gulp, and hope your insurance agent doesn’t find out about it. It was also used to figure out the optimum organization of slot machines on the floor of a casino. My mother worked as a school principal in the ’90s and used a spreadsheet to keep track of students and classes. One day she called me up and said, “Hey, did you know this thing can also do calculations?” There was a time when even businesspeople didn’t always know that. Some people kept cheap calculators glued to their computers to add or subtract or whatever before entering the values into the cells!
Nintendo probably doesn’t have to fight Excel for mind-share among 10-year-olds, but today people are exposed to spreadsheets quite early.
Yes, spreadsheets are now being taught, both in terms of the way of thinking they represent and how to actually use them, from a young age. They don’t necessarily teach it in the guise of an electronic spreadsheet, but they say, “Here’s a column and a row, we do this and try that, what formula would we use for this one?” and so on.
That would seem to argue for spreadsheets always being with us.
It’s a way of working that seems very natural, although for certain uses, such as databases, other products have replaced it, and may continue to replace it. For some financial applications, users may accept more restrictions because of what the software that imposes them can do.
In our day, when we first started computerizing things, we tried different ways of theorizing, and new ideas and new tools came about; constantly some things kept on getting replaced. So we all assumed that whatever we did would be replaced by a new thing later on. That was the Y2K problem. None of us believed that our code wouldn’t be replaced in a few years. The pace of change was such that who would have thought that their programs would be around in the year 2000, to the point where decisions about how to handle the calendar actually posed a problem? We would have considered it egotistical to regard our code as analogous to bridges and waterways, meant to last 50 years or more, but in many cases, it played out that way. Looking at the spreadsheet, we never thought it would last this long. After all, we were still inventing new tools.
It’s like walking across the United States and you come to the Grand Canyon and you say, “Oh, this is really cool! I wonder what other cool things we’re going to find!” But there’s only one Grand Canyon.
A Thousand Times Better, and Counting
The electronic spreadsheet may be ubiquitous—Microsoft estimates there are 400 million users of its Office product worldwide, which includes the now-dominant Excel spreadsheet along with Word and other desktop applications—but when Dan Bricklin and Bob Frankston were developing VisiCalc, it was a decidedly regional affair. Mitch Kapor was among the Boston-area technophiles who got an early look, and he recalls, “I was just blown away. It was a thousand times better than the graphical and statistics program I was developing at the time.”
Kapor couldn’t have known it then, but he would go on to become nearly as identified with the electronic spreadsheet as its creators, thanks to his role as co-founder of Lotus Development Corp., the company that made the Lotus 1-2-3 spreadsheet a corporate staple. While the Lotus version added macros, graphing, and other improvements to the spreadsheet, Kapor maintains that it has not fundamentally changed in the past 25 years. But, he adds, that doesn’t mean it can’t. Joe Krawczak of Microsoft says the company plans continual enhancements to Excel, many of them designed to make it easier to get data into spreadsheets and to share them. In fact, while its rows and columns may comprise thousands of cells, the spreadsheet itself is still, in Kapor’s words, “the equivalent of a single-cell organism, with plenty of room to evolve.” —S.L.