Empirically Derived Use Cases for Software Analytics

Abstract

[Background] Software engineering activities provide large volumes of data that software analytics tools can use to support decision-making. However, adopting such tools depends on the usefulness of the information provided regarding the needs of practitioners. While the needs of developers have been well-researched, the needs of managers are not getting as much attention. [Aims] This study provides an in-depth analysis of the needs of software practitioners involved in managerial decision-making from one organization that performs research, development, and innovation projects with industry partners. [Method] We identified and represented such needs as use cases by interviewing people in leadership positions and analyzing the collected data using Grounded Theory coding techniques, i.e., open and selective coding. [Results] Our analysis resulted in 19 software analytics use cases which we classified into four dimensions: quality, people, project management, and knowledge management. The use cases in the quality and project management dimensions were the most mentioned ones. [Conclusions] Although our results are particularly relevant to organizations similar to the one described herein, they aim to serve as input for implementing new analytics solutions by practitioners and researchers.