Christopher Drum's Improv retrospective has people asking me about Lotus Improv again. It turns out I've been writing about it on this blog, on and off, since 2004. So here they all are in one place.
Where the idea came from. The newest post tells the oldest part of the story. In Lotus Improv Pre-History I describe reading a pile of "how to run your inventory in 1-2-3" books and noticing the same formula, copied by hand across groups of columns, over and over. Write it once and let the program apply it. I called that the semantic spreadsheet. Showing the same data with its dimensions flipped came next, and that became the pivot table.
Why it didn't succeed. In 2004 Adam Bosworth wrote that Improv "failed utterly, not because it failed in its ambitions but because it succeeded." In Why Improv didn't succeed I agreed with him more than I expected to. In setting out to improve on spreadsheets, Improv lost what makes a spreadsheet a spreadsheet. I also wondered whether the Innovator's Dilemma applied, and whether we should have started in a niche where all that structure was a feature. A few days later, in Part Deux, Peter Murray of Quantrix wrote in to say the Improv model was not ancient history at all.
Disrupting yourself. Eating your young is about the gap between talk and launch. While we were building it, far from the sales force, "throw out the rule book" was easy to say. Once it was time to sell it, Lotus was a lot less eager to disrupt 1-2-3.
Other people noticing. There's a surprisingly accurate Wikipedia article, which I had nothing to do with. A book on pivot tables credits Improv with the idea. PC Magazine's 25th anniversary issue listed it among the big deals of 1993, with a bad photo of me. And Chris Herot, who built Javelin, pointed out how much Improv borrowed from it.
The problem it set out to fix. In 2013 I came across a study saying 88% of spreadsheets have errors. We started Improv around 1987 because companies were afraid to touch the spreadsheets they ran their business on. Twenty-five years later, same story.
So my lesson, thirty-some years on: a better idea doesn't get you very far if you sell it as the old thing. Improv wasn't a spreadsheet, and we kept telling people it was.