Respondents to the ECAR study by Goldstein said that they have not had a lot of success in reducing the presence of shadow databases. Shadow databases are "silo" information systems maintained by separate functional units in the organization (or even within the same unit) which contain the same types of information, or perform similar functions. The problem is, usually they do NOT contain the same information. Shadow databases proliferate when there is weak oversight of the institution at the highest levels of administration, when little communication exists between units, and when individuals are rewarded not for sharing information but for hoarding it. A perfect description of many institutions in the world of higher education.
The problem with shadow databases is that they represent wasted time, effort, and money. Not just by wasting the opportunity to learn from others' information, but sometimes even in the money put toward purchasing redundant systems. But they do serve an important sociological function in the workplace: preserving the autonomy of a unit within a large organizational structure.
Institutions with "performance dashboards" as part of their analytics technology report the best outcomes for eliminating shadow databases. No doubt this is because the dashboard tools are very attractive analysis tools for almost every organizational unit, and this makes units want to opt in. Just as likely, however, is that these systems suddenly make it glaringly obvious whose information is missing, thereby providing some pressure from peers and higher-ups to participate.
Eliminating shadow databases is indeed a business improvement that the faint-hearted should not attempt.