L is for Lotus Development. If there is a lesson from the tale of Lotus, it is that in the world of software it is not enough to get it right once: you have to get it right again and again, or you will get bought up by someone else and end up a footnote in history books.

Lotus Development was founded in 1982, the brainchild of Mitch Kapor and Philippe Kahn. Their brilliant idea: since the top PC programs of the time were wordprocessors, spreadsheets and databases for the IBM PC, how about clumping the three together into one package? Like all the best ideas, it was so obvious that nobody had done it, but everyone wanted it.

1-2-3, as the program was called, was an enormous success, selling more than 14 million copies. Because you could share data between the programs, you could do work in minutes that previously would have taken hours. It became the rationale for buying PCs and computerising entire offices. Microsoft's MS-DOS might have been the bread on which the PC revolution was built (and the IBM PC the rather bulky toaster), but 1-2-3 was the butter, jam and marmalade.

Lotus got stuck trying to find a follow-up to 1-2-3, while rivals moved into other fields. Eventually, in 1990, it came up with Notes - a product designed for the "virtual corporation" that allowed people out of the office to interrogate and update databases via modems.

Notes was a success, but it was still surprising when last year IBM bought Lotus for $3.5bn. Ever since, everyone has been waiting for something that would show that it made a clever buy. Meanwhile, Microsoft is still independent, and growing, proving that to survive in this business you must get it right and keep on getting it right.