Published Sep 3, 2019

SE-Radio-Episode-273-Steve-McConnell-on-Software-Estimation

Steve McConnell delves into the complexities of software estimation, highlighting the importance of early risk identification, effective use of estimation techniques like planning poker and story points, and the alignment of Scrum practices to manage uncertainties and changes in project scope. The episode emphasizes the need for clear communication between business and technical teams to differentiate estimates, commitments, and targets effectively.
Episode Highlights
Software Engineering Radio - the podcast for professional software developers logo

Popular Clips

Episode Highlights

  • Risk Identification

    Identifying risks early in a project is crucial for effective risk management. emphasizes that Scrum practices, with their short iterations, naturally lead to early and frequent risk identification. This proactive approach allows teams to address risks with higher leverage options, reducing significant inherent risks and sources of variability in software projects 1.

    If we can identify risks early rather than late, we typically have way higher leverage options for addressing those risks.

    ---

    adds that agile methods, like Scrum, help mitigate intrinsic risks by requiring deliverables every two weeks, which must be accepted by stakeholders, thus reducing the risk of incomplete work 2.

       

    Uncertainty Management

    The cone of uncertainty is a critical concept in project estimation, highlighting the variability and unpredictability in software projects. explains that effective management involves attacking the highest sources of variability first, reducing uncertainty as the project progresses 3.

    The cone of uncertainty is a way of describing what can happen on a healthy project.

    ---

    However, many teams fail to manage this effectively, often leaving the hardest problems for last, which complicates estimation and project control. McConnell also notes that early project estimates are often unreliable due to numerous unknowns, emphasizing the need for rough sizing based on historical data to set realistic expectations 4.

Related Episodes