A Methodology for Improving Business Process Performance through Positive Deviance

The importance of process improvement and the role that best practice reference models play in the achievement of process improvement are both well recognized. Best practice reference models are generally created by experts who are external to the organisation. However, best practices can be implicitly derived from the work practices of actual workers within the organisation, especially when there is opportunity for variance within the work, i.e. there may be different approaches to achieve the same process goal. In this paper, the authors propose to support improvement of process performance intrinsically by utilizing the experiences and knowledge of business process users to inform and improve the current practices. The proposed methodology is inspired by the theory of positive deviance. By utilizing a multiple criteria decision making approach and Shannon’s entropy method of information theory in determining useful information from uncertain data within execution log of business process, the authors are able to define the “best” and most suitable previous practices as a recommendation that fits with the current competence/experience levels of individuals. The authors demonstrate that the proposed method is capable to generate meaningful recommendations from large data sets and effectively facilitating learning within organisation leading to process performance improvement.