The story of the evolution of spreadsheets is well-known, though rarely studied. Spreadsheet evolution illustrates the patterns of software evolution. It shows the power of those patterns; if the leaders of Apple, Lotus and Visicalc had known and acted on the predictive power of those patterns, Office would not have been a Microsoft product, but one of theirs.
As a reminder, here are the basic patterns of software evolution:
Software is created for a new platform.
Software grows more capable on the platform with increasing abstraction.
The automation provided grows deeper.
The breadth of automation grows wider, doing more things.
The Spreadsheet products in evolutionary terms
The first popular spreadsheet program appeared on the then-new Apple computer. It was Visicalc. In terms of the stages of evolution, it was this:
- Emergence on a new platform: Conceptual Breakthrough
- Development on a platform: Basic product
- Automation depth: Recorder
- Automation breadth: Point product
If the platform became dominant and remained stable, we would expect competition by increasing the depth of automation on the Apple, increasing the breadth of automation e.g. by creating a collection of related products, or by moving along the same-platform development dimension. Instead, a powerful and popular new platform emerged, the PC. Visicalc themselves should have become the leaders in spreadsheets for the new platform, but as so often happens, the emergence of the new platform is an opportunity for a new, focused company to do it.
The company that won was Lotus, with their 1-2-3 product.
- Old platform: Apple I/II
- Old functionality: Spreadsheet, like Visicalc
- New platform: IBM PC/DOS and clones
- New functionality: Spreadsheet, much like Visicalc, along with some hype of other applications
- Outcome: They won big time
All the other categories (Basic product, etc.) remained the same.
Meanwhile, back in the land of Apple, the Mac came out, the first practical GUI on a personal computer platform. You would think this would be the big chance for Apple themselves to jump into the business, or for Lotus to convert their category-killing program. And of course, it was their big chance, but they blew it. Instead, Microsoft of all unlikely companies, jumped in.
- Old platform: Apple I/II, PC/DOS
- Old functionality: Character mode spreadsheet, like Visicalc and Lotus 1-2-3
- New platform: Apple Mac GUI
- New functionality: Excel Spreadsheet, but with attractive fonts and graphics
- Outcome: They won on the Mac
It wasn’t easy to get a spreadsheet right for GUI, but Microsoft kept plugging away and really took advantage of the graphics and the mouse.
At the same time, Microsoft was plugging away at their own clone of the Mac operating system for the PC, Windows, which they announced in 1975 and finally came out with a usable release, 3.1, in 1982 (seven loo-oo-oong years later – but that’s another story). Here we have a GUI platform for the PC, Lotus’ home base. The GUI had shown its attractiveness elsewhere. There was a GUI backer who was clearly going to see it through. You would think that the savvy business executive who was in charge of Lotus at the time, surrounded by a heavy-weight group of Cambridge thinkers, would have been all over this. Didn’t happen. While Apple built the GUI but failed to build the spreadsheet for it, and while Lotus built the spreadsheet but failed to convert it to a GUI on their own dominant platform, Microsoft built both the GUI and the spreadsheet.
- Old platform: Apple Mac, PC/DOS
- Old functionality: Mac GUI spreadsheet; PC/DOS character spreadsheet
- New platform: PC Windows
- New functionality: Same as on Mac
- Outcome: Huge win/win: PC/Windows helped Excel, Excel helped Windows
At the time, Lotus had much more revenue than Microsoft, and seemed reluctant to rewrite its industry-leading spreadsheet for Microsoft’s Windows, which was seen as a move that would help the success of that platform, and thus help Lotus’ competitor. But once it was clear that Windows was here to stay, Lotus finally did the re-write, and got 1-2-3 onto Windows. So now 1-2-3 and Excel were competing as point products on the same technology platform.
But now Lotus had another problem: 1-2-3 was a point product (I know, I know, they said they had three applications, thus the “1-2-3,” but who used the “2-3”?), and by this time, Microsoft had an actual product suite: Microsoft Office, combining spreadsheet with word processing and presentations via OLE and drag-and-drop. On the movement from point product to product collection to product suite, Microsoft was consistently a step ahead, and thus had more market pull than Lotus. While Lotus was plugging the virtues of 1-2-3 vs. Excel, Microsoft was promoting the value of its product collection. Then, while Lotus was promoting the value of its acquired collection of office products, Microsoft had a real product suite. The end result, we know today, is the complete dominance of the Microsoft Office product suite.
Conclusion
The evolution of the spreadsheet fits nicely into the general patterns of software evolution I have described. If strategists and company leaders had been aware of these patterns and acted on their proven predictive power, the outcome of the spreadsheet wars would have turned out differently. Leaders today would be well-advised to understand and apply the patterns of software evolution.