A lot of enterprise software was changed or spawned during the 1990s and the first decade of this century based on the success of the Toyota Manufacturing (or Production) System. What is less well documented but clearly linked is that enterprise software development itself also changed during the same period for the same reason.
The terms agile or waterfall tend to be used rather than any mention of Toyota (although I am aware of one source that talks about kanban software development) but the concepts of managing by walking around, standardized work, the worker stopping the assembly line, the extended supplier community supplying parts just in time and so forth sounds a lot like community forums, rapid application development, "The Cathedral and the Bazaar," re-usable objects from a repository and similar ideas that permeate the world of programming these days.
That leads to two questions.
- Did the philosophy behind the Toyota Manufacturing/Production System lead to what now looks to be Toyota's quality-control implosion?
- If so, and if 21st century enterprise software development follows the same "production" philosophy, is a quality-control implosion likely in enterprise software development in the near future?
(These questions are unrelated to the fact that embedded software developed in some Toyota Way may actually be the culprit behind the quality problem, as opposed to oversized floor mats or some 99-cent gas pedal that is going to cost $500 per vehicle to replace.)
Consider the ripple effects to enterprise software investing if the answer to the two questions above is "Yes." In my opinion, it will mean that a handful of related enterprise-software-market trends will be accelerated:
- The end of the home-spinning-wheel era of software development and finally the completion of the 20-year-old concept of the software factory
- More mergers among software, systems and IT services suppliers (e.g. Perot into Dell (DELL), Sun into Oracle (ORCL)) such that there will be no distinction among types of IT companies; and of course there will be fewer, bigger of the remaining category: IT companies able to afford sofware factories (just as fewer and fewer companies could afford to build and run fab plants during the 90s)
- The movement to most information processing to the "cloud," completing the 50-year return to utility computing.
There are a lot of other similar back to the future implications, all just a matter of time (which all mean there is a time to buy and a time to sell in an increasing number of publicly traded enterprise software suppliers).
-- Dennis Byron