Quantifying the Effect of U sing Kanban vs . Scrum : A Case Study

Many claims about the usefulness of various processes or methods, such as Scrum and Kanban, have been stated in agile and lean software communities. However, these claims are rarely supported by objective data or empirical investigations. In contrast, this article aims to demonstrate that the effect of processes or methods (here: Scrum versus Kanban) can be evaluated and compared on the basis of objective data. We analyzed data of more than 12,000 work items collected over the years 2009-2011 in a medium-sized software company. The company used Scrum from 2007 to autumn 2010, at which point they changed to Kanban. By using Kanban instead of Scrum, the company almost halved the lead time, reduced the number of weighted bugs by 10%, improved productivity by 21% for PBIs, and reduced productivity by 11% for bugs. Consequently, Kanban seems to outperform Scrum in this company. However, the results should be interpreted with caution because the use of Kanban succeeded the use of Scrum. To acquire more knowledge about the performance of different agile or lean methods, scholars should conduct similar studies in different organizations in different application domains and with people of different cultures and competences.