Doctoral dissertation (article) (G5)
On top management support for software cost estimation
List of Authors: Rahikkala Jurka
Publisher: Turku Centre for Computer Science
Place: Turku
Publication year: 2019
eISBN: 978-952-12-3835-2
URL: http://urn.fi/URN:ISBN:978-952-12-3835-2
Self-archived copy’s web address: http://urn.fi/URN:ISBN:978-952-12-3835-2
Inaccurate software cost estimates continue causing project overruns and hurting firms’ economy. This thesis addresses the problem by focusing on top management role in applying estimation methodologies successfully in organisations. The research questions are 1) How does top management support software cost estimation, and 2) What are the impacts of top management support for creating a good cost estimate for a software project?
Three empirical studies, one quantitative and two qualitative, were conducted to address the research questions. The studies identified practices, through which top management is involved in cost estimation, and collected evidence on the impact of practices on estimation success. The quantitative study is based on views of 114 Finnish software professionals, and the quantitative studies are based on in-depth findings from three Finnish software producing companies and projects.
The results show that top management support for estimation is mostly indirect. Management focuses on creating a successful environment for estimation instead of hands-on participation.The key factors of top management support include adequate resources, demonstrating the importance of estimation and seeking realism. This indirect role is enough for successful estimation.On the other hand, the results provide evidence that top management may negatively impact estimation. For example, unclear expectations may cause the project team to aim for the wrong outcome, expressed expectations may bias estimation and interpreting estimates as commitments may decrease estimators’ motivation and cause them to give high estimates.
The practical implication is that top management should avoid direct participation in software estimation and focus on sustaining a supportive and unbiased environment. By doing this, many projects should be able to avoid failures hurting firms’ competitiveness. From the research perspective, the results provide evidence that people-related perspectives are an important factor in software estimation, implying that a shifting focus from methodologies toward managerial topics is justified.