The Competitive Crash in Large-Scale Commercial Computing

We examine the factors underlying buyer demand for large Information Technology solutions in order to understand the competitive crash in large scale commercial computing. We examine individual buyer data from two periods. The first is in the mid 1980's, late in the period of a mature and stable large-systems market. The other period is in the early 1990's, very early in the diffusion of a new, competitive technology, client/server, when many buyers chose to wait for the new technology to mature. We clarify the implications of different theories of the competitive crash and then test them. The most popular theories are far wrong, while the correct view emphasizes the 'internal' adjustment costs to organizations making IT investments. Understanding buyer behavior not only illuminates the competitive crash, but also the factors underlying the slow realization of the social gains to Information Technology in large complex applications more generally.