I’ve just finished getting through the Kindle version of this book and I’m glad to say I enjoyed it. Even though I’d read a lot of the points before, the purpose of Craig Larman’s book is approach this from a business viewpoint and why an organization should move to Agile.
Larman is a known expert and advocate for object-oriented technologies (he authored Applying UML and Patterns, a best-selling text on object-oriented analysis and design) and development, and iterative and Agile development methods. He is also the Chief Scientist at Valtech. His role has seen him lead a number of companies to adopt iterative and Agile methods. Needless to say, he is not an unbiased or disinterested party in the Waterfall vs. Agile debate.
Setting out from page one with the purpose to equip the reader with the evidence needed to convince a skeptical management that Agile is no passing fad, he makes no bones about the fact it should be considered seriously. Indeed, there is a whole chapter is dedicated to the research, historical evidence, and other evidence that incremental iterative development works, whereas Waterfall can be risky.
The book starts with a background the Agile movement before diving into the evidence to support Agile. He focuses on 6 key areas:
- Research evidence showing iterative and evolutionary development produces better results (measured in level of risk, number of defects, and level of productivity) when compared to Waterfall projects.
- Early large project evidence where major and life critical projects have been developed using Agile methods.
- Standards-body evidence such as where the US DoD, frustrated by the failure of Waterfall projects, went from mandating Waterfall to preferring iterative methodologies.
- Expert thought leader evidence from prominent software engineers who recommend avoiding Waterfall and moving to iterative approaches.
- Business case around specific numbers that back up the benefits of an iterative approach (for example, lower failure rates with iterative projects whereas for Waterfall they had 23% of projects, averaging $1.1 million USD, fail.
- Waterfall problems and specifically how the original Royce paper was misinterpreted (I covered Royce’s paper in my Waterfall post).
- Why Waterfall is still promoted, with several specific reasons why.
Larman then goes into several methodologies in detail (specifically Scrum, Extreme Programming, Unified Process, and Evo) before wrapping up with a chapter on putting Agile into practice.
In summary, the book is not an in-depth view into a single methodology, but provides a strong business case for the move to Agility using four detailed examples to compare and contrast the various methods that a manager looking into Agile for the first time may come across. If you need to make a case to management to move to Agile, this book will give you a raft of facts and figures to make for a strong one.