Managing knowledge in the dark: an empirical study of the reliability of capability evaluations

If knowledge is to be managed and transferred, it is essential that members of organizations know and agree on where capabilities reside. Few studies, however, have examined the difficulties of evaluating capabilities in large firms. This paper reports an in-depth empirical study of capabilities central to knowledge management efforts in large leading multinational companies. The results show that evaluation of these capabilities is a complex task. The median interrater correlation for capabilities designated as strategic by top management is only 0.28. Analysis of the determinants of reliability show that the difference in evaluations is largest for subsidiaries managers know less about, for younger subsidiaries, and for subsidiaries in less important markets. The results of our empirical study have important implications for creating, retaining, and transferring knowledge in organizations.