I just finished reading “Agile Project Management with Scrum”, but the Scrumm man himself (Ken Schwaber), so thought I would share a quick review:
Book review - Agile Project Management with Scrum / Ken Schwaber
This book uses a series of case studies to examine solutions to various complications that can arise in a Scrum (or any other) project. The case study format is informative and reasonably well written, although it doesn't have the rigor or authority of more quantitative studies.
I would recommend the book for any project managers (/Scrum Masters) that are involved in a Scrum project; it's a short enough read that they should get value from reading it before or at the start of a new project. The book also has a quick overview of Scrumm – a good reminder if you aren’t that familiar with the process.
Examples focus on the different roles and artefacts of Scrumm, focussing on guidance for each area. One section I found particularly good was guidance for scaling Scrum projects (using a timeboxed "Staging" process to define the required non-functional scalability requirements as requirements).
Some of the examples given really demonstrate the power of iterative methodologies and focussing on delivering business value. One particularly memorable example (p.19) was introducing Scrum on a project to automate import of land ownership changes from various state governments; the project had failed twice before.
By focussing on delivering "potentially releasable" software in the first iteration (4 weeks), focussing on hte highest value feature. The business then decided to actually release it (in a short, 2 week, second iteration), reducing their workload by 40% after only 6 weeks.
The book is relatively thin (around 160 pages), which always gets high points to me, so it is a relatively quick read. Although tied to the Scrum method, it does have some longevity (compared to books around specific technologies). Some of the examples are useful for general iterative development even if you aren't doing Scrum or even if you are using a heavier (non-agile) iterative process.
One thing that constantly irked me was the author confusing Gantt charts and something that is like a PERT network diagram (p.88), calling it a "Gantt report"; I know its picky, but it just really annoyed me and detracted from what is otherwise a reasonably good book.
I would give it 3.5 / 5 as a general software project management book, compared to classics such as Peopleware or the Mythical Man Month. As a book about Scrum, however, I would give it 4 / 5.