The Crisis in Software: The Wrong Process Produces the Wrong Results
Ken Schwaber, Jeff Sutherland · 2012
This chapter discusses the crisis in software, where the wrong process leads to wrong results. Business, governmental, or non-profit organizations need to be able to create value by building, customizing, and using software. Without software, the ability to achieve goals as a business leader is inherently limited in fact not impossible. But despite this need, software development has historically been an unreliable, costly, error-prone endeavor. The chances that a software project will be successful are not good. The software industry has failed one by being slow, expensive, and unpredictable. If software weren't so important, one would probably stop investing in software altogether. The organizations have typically been struggling to increase the success rates for their software projects. They seek help because they fear that their software organizations are spinning out of control. Their existing process has failed them, and they do not know of an alternative. Their problems with software development create a tremendous amount of waste for their organization, yet they persist because they are dependent on software to remain competitive. Software development has in the past been prone to failure. The root cause of this failure is the use of predictive processes for complex work. When one shifts to Scrum, an empirical process, the software project success rate increases dramatically. It is possible to get software features ready to use in a month or less but not hundreds and thousands all together. A software product may still be big, but it can still be built in small pieces, one by one, in a month at a time avoiding wrong results.