Accelerate: The Science of Lean Software and DevOps: Building and Scaling High Performing Technology Organizations Paperback – Illustrated, March 27, 2018

This book was excellent, I read it twice before reviewing. They scienced the crap out of DevOps and go to painstaking lengths to identify what can and should be measured, and what those predict. Very helpful if you are looking to baseline your DevOps capability and identify areas for improvement around key outcomes.Fair warning though, the book doesn't go into detail on how the capabilities identified/measured work together, the correct combinations, or proper order. Recommend picking up Lean Enterprise and Continuous Delivery by Jez for that. So it's not that the authors are naive to this, it's just not the intent of this book. On the contrary I'd say they are very aware of what can be measured and what would be speculation--and this book is scoped to what they could measure.With that in mind, it actually places appropriate limitations on the conclusions you can draw from their results--an atypical (but welcome) admission they make early on. They can only measure outcomes, not intents. So they show behaviors and how they correlate to outcomes, but warn the reader to understand the why behind those behaviors before trying to adopt them blindly. Basically a nice way of saying, "Don't be a #cargocult."In Lean Enterprise, Jez and his co-authors make several recommendations on combinations and order, as well as the why behind each. I appreciate that this book only focuses on what can be measured--it provides a good baseline from which to look at my own organization's practices. But it does assume that the reader is familiar with DevOps and isn't looking to understand the why or the how.I highly recommend this book.

Accelerate will convince you that modern agile/devops development practices are worth investing in and will bring you business benefits. At least, that is the goal of the book. It explores survey results from 3 years of DevOps survey, explain why they are trustworthy and relevant and what you can learn from them. It does so in a very convincing way if I may say so. I personally experienced most of the promoted practices to be useful and therefore probably didn't need convincing. If I did, this book might have been able to do so.The DevOps survey is an industry survey originally done by Puppet Labs for exploring Continuous Delivery and DevOps practices in the industry. The first DevOps survey was in 2014 and the book takes 3 years of survey results (3 surveys) and shares the results and the conclusions of these results. The book consists of three parts: (1) What we found, (2) The Research, and (3) Transformation.The first part shares the results and conclusions of the DevOps survey. Good development and continuous delivery practices result in less stress, better quality, and better business results. This part summarizes different practices and how they correlated with improved business success. I felt most of the practices were not controversial (for someone with an agile background) although there were some exceptions (how far should you go in not standardizing tools) and areas not covered. Especially the area of organizational and team structure was not covered and, at times, the book suggested traditional organizations and traditional role divisions. This was unfortunate as it would have been interesting inclusions... but not covered well in this book.I actually enjoyed the second part of the book, which had nothing to do with software development but explains the different research methods and practices applied. It explains different data collection strategies and why a survey was the right strategy for the questions the authors were asking. One skepticism I had (still have) is that the selected target population (people familiar with DevOps) causes a self-selection bias and therefore invalidates the findings when extrapolating to the entire industry. The authors, unfortunately, didn't discuss that much, but it did come up with arguments on why they should restrict the target population to people familiar with DevOps. The arguments were good... though not fully convinced me. Still, I found part 2 unusual and interesting.Part 3, transformation, was small and not written by the authors. Instead it provided a case study of lean management practices by Steve and Karen Whitley Bell. The case study was from ING Netherlands. Although I enjoyed the case study, I did wonder at times why it was included as it didn't actually talk about the majority of the practices of the book. It mostly focused on Lean Management and Lean Transformation practices. Nevertheless, I enjoyed reading the case study.All in all, Accelerate was an enjoyable little book. It didn't provide huge new insights to me, which was not the intention of the book. The intend was to share evidence (science) that some existing modern practices actually work. In that, the book succeeded. I would not recommend the book to people who want to understand these modern practices in-depth, for that, this is the wrong book. I would very much recommend the book for people who want to understand (and be convinced) that these modern DevOps/Development/Agile practices can have a positive effect on your business... and they are worth investing time and resources in. Good book, recommended, 4 stars.

Many of the recommendations in the book seem like common sense. However the authors have taken the time to explain how they came to the conclusion and presented a case study of how leaders help transform teams. It’s a good read for anyone trying to set standards for tech teams at scale and transform organizations.

Even in 2024, Accelerate remains a must-read for improving software delivery performance. Seven years after its release, the book’s lessons are as relevant as ever, thanks to its practical advice and real-world data. While the first part is especially impactful, the entire book offers valuable insights that apply seamlessly to modern AI-driven tools and systems.One of the book’s greatest strengths is its focus on achieving meaningful outcomes rather than merely checking boxes. It makes a compelling argument against maturity models, stating:> “Maturity models are not the appropriate tool to use or mindset to have. Instead, shifting to a capabilities model of measurement is essential for organizations wanting to accelerate software delivery.”The authors emphasize measuring what truly matters, such as delivery lead time, deployment frequency, time to restore service, and change fail rate. These metrics ensure teams focus on results, not unnecessary busywork. As they put it:> “A successful measure of performance should have two key characteristics. First, it should focus on a global outcome to ensure teams aren’t pitted against each other.. (Second,) it shouldn’t reward people for putting in large amounts of busywork that doesn’t actually help achieve organizational goals.”Overall, Accelerate is an exceptional guide for improving software delivery and team performance. Its ideas are simple to grasp, practical to implement, and enduringly relevant. For anyone working in tech or leading teams, this book is just as valuable today as it was when it first came out.