Daniel Galorath, President and CEO

For over 30 years, Daniel D. Galorath has been solving management, costing, systems and software problems for both information technology and embedded systems. One of Dan’s strengths is reorganizing troubled software projects, assessing progress applying methodology and plans for completion and estimated cost to complete. Dan’s company, Galorath Incorporated, develops the SEER™ applications for software, hardware, electronics & systems, IT and manufacturing cost, schedule, risk analysis and management decision support. He is one of the principal developers of the SEER-SEM™ software evaluation model. Dan’s book, Software Sizing, Estimation, and Risk Management was published in 2006. Follow Dan’s blog at http://galorath.com/.

演讲题目： 敏捷软件项目估算的偏差和缓解（Estimation Bias and Mitigation with Agile Estimate Considerations）

Agile software development is taking the world by storm. There is much good in Agile, but also a number of dangers. One issue within the Agile community is the insistence that estimates are not needed. Combine this with the bias that comes from less than rigorous estimation practices, and you have an opportunity and a risk. On the opportunity side, many Agile programs are more productive, playing down documentation and using working software as the major criterion for success. These work well with smaller systems and IT systems where the potential user is engaged on a daily basis, but can create significant risks for larger or embedded systems. Hybrid Agile can help mitigate the risks. And viable estimation processes can give management the insight they need up front and provide the basis for earned value management on larger systems.

This paper will raise awareness of estimation bias and discuss Nobel Prize winning work on estimation bias to substantiate the need for proper estimation in both Agile and traditional development environments. There is much evidence that viable estimates can make projects successful, make outsourcing more cost effective, and help businesses make the most informed decisions irrespective of the development methodology used. Stakeholders care about cost and schedule irrespective of whether the project is Agile, traditional, or hybrid.