Information Technology Reference
In-Depth Information
cycles and a higher emphasis on quality are encouraged) and engineering (e.g.,
design-for-manufacturability is encouraged).
Given the enormity of the potential benefits extending across multiple func-
tions and indeed the entire enterprise, one might expect that manufacturing firms
would have been especially quick to adopt and deploy this technology. However, the
actual history of CAD adoption followed a much different story line. The technol-
ogy was indeed rapidly acquired by manufacturing firms, but many years elapsed
before it was actually utilized in a way consistent with the vision of the technol-
ogy's designers. Liker et al. report that as late as in 1992, a decade after CAD was
introduced, “true CAD/CAM [utilization was] still quite rare” (Liker, Fleischer, &
Arnsdorf, 1992).
A variety of explanations could account for the slow deployment of CAD, such
as technological immaturity, the difficulty of organizational learning, and incen-
tive conflicts. However, the logic of complementarities itself provides an additional
compelling explanation: If the majority of CAD's benefits only arise when the tech-
nology is combined into a complementary system of elements, this fact would
actually serve to magnify the ill-effects of technological immaturity, learning bar-
riers, local incentive conflicts, etc. An immature technology tends to have “bugs”
(features that are missing, underdeveloped, or just do not work as they should). If
benefits are not materializing, how does an organization sort out which problems are
due to “bugs” in the technology, or “bugs” in the design of the surrounding organiza-
tion? If the technology itself is hard to understand due to knowledge barriers, it will
be that much more difficult to anticipate the best configuration of complementary
organizational elements to build around it. If the technology poses incentive con-
flicts, that will make it more difficult to rally the whole organization around the need
to make complementary organizational changes. Thus, in what might be seen as a
supreme irony, complementarities not only magnify the beneficial effects of innova-
tion investment when things go favorably, but may well make it less likely that things
will go favorably by magnifying the effects of typical implementation barriers.
2.3.1 Complementarities-Based Studies in Innovation
and IT Business Value Research
As the CAD example shows, IT investment entwines with organizational innovation
and business value in a manner consistent with the logic of complementarities. Thus,
it is not surprising that complementarities have been receiving increasing attention
from both innovation scholars and IT business value researchers. In this section,
we step back to formalize the logic of complementarities and briefly survey some
important empirical work.
According to Milgrom and Roberts, two activities are “Edgeworth” complements
if “doing (more of) one thing increases the returns to doing (more of) the others”
(1995: 181). A necessary condition for the existence of complementarities is that
the effects of two variables (A, B) on a third variable (C) be supermodular ; that is,
Search WWH ::




Custom Search