Accelerate

Book Review: Accelerate by Nicole Forsgren, Jez Humble, and Gene Kim

Introduction

In the rapidly evolving landscape of software development, organisations strive to deliver high-quality software swiftly and reliably. Accelerate: The Science of Lean Software and DevOps offers a research-backed exploration into the practices that distinguish high-performing technology organisations.

This review dives into the book's core insights, emphasising strategic, action-biased takeaways, with references to complementary resources for leaders, practitioners, and teams aiming to deliver better software, sooner, safer, and happier.

The Essence of Accelerate

Accelerate is grounded in four years of rigorous research conducted by the authors and the DORA (DevOps Research and Assessment) team. Analysing data from over 23,000 professionals worldwide, the authors aim to identify the capabilities that lead to both superior software delivery and organisational performance.

The book is structured into three parts:

  1. Research findings and key takeaways
  2. Scientific methodology and statistical modelling
  3. Real-world application through case studies

Key Findings and Actionable Insights

1. Four Key Metrics of Software Delivery Performance

The research identifies four critical performance metrics:

  • Deployment Frequency – How often code is deployed to production.
  • Lead Time for Changes – Time from code commit to deployment.
  • Change Failure Rate – The percentage of deployments causing production failures.
  • Time to Restore Service – Time to recover from failures.

ℹ️ Action:

Start by measuring these four metrics. These indicators offer a clear and manageable view of your delivery performance. Adopt tools such as DORA metrics dashboards, integrated with CI/CD pipelines, to track and improve performance.

Supporting Resource: Google Cloud DORA Metrics

2. Technical Capabilities that Drive Performance

Accelerate surfaces 24 key capabilities that drive software delivery performance. Grouped into the following domains:

  • Continuous Delivery: Version control, deployment automation, continuous integration, trunk-based development
  • Architecture: Loosely coupled systems, team empowerment
  • Product & Process: Customer feedback loops, working in small batches, team experimentation
  • Lean Management & Monitoring: WIP limits, lightweight change approvals, proactive monitoring
  • Culture: Generative culture, psychological safety, transformational leadership

ℹ️ Action:

Run a capability assessment using these categories. Identify your strongest and weakest areas. Then, incrementally adopt and improve the capabilities most aligned with your strategic goals.

Supporting Resource: The DevOps Handbook by Gene Kim et al. expands on the implementation of these capabilities.

3. The Role of DevOps in Organisational Success

High-performing teams:

  • Deploy on-demand with high frequency
  • Recover from failure quickly
  • Maintain a low change failure rate
  • Achieve short lead times

These performance advantages also correlate with higher job satisfaction, lower burnout, and stronger business outcomes.

ℹ️ Action:

Invest in DevOps training, build internal capability through communities of practice, and embed DevOps into your engineering operating model.

Supporting Resource: Puppet’s State of DevOps Reports provide longitudinal insight into DevOps maturity and outcomes.

4. Cultural Factors Are as Important as Technical Practices

Westrum’s organisational culture typology is referenced heavily:

  • Pathological (power-oriented)
  • Bureaucratic (rule-oriented)
  • Generative (performance-oriented)

Generative cultures – those that emphasise information flow, learning, and cooperation – consistently outperform the rest.

ℹ️ Action:

Adopt cultural interventions such as blameless retrospectives, skip-level feedback loops, and leadership visibility.

Supporting Resource: Team Topologies by Skelton & Pais aligns well with these cultural dynamics.

5. Scientific Rigor and Statistical Validation

Unlike anecdotal frameworks, Accelerate is built on peer-reviewed research and validated statistical models (e.g. structural equation modelling). It demonstrates causation, not just correlation.

ℹ️ Action:

Use this data to influence stakeholders and justify transformation investments. Ground your DevOps strategy in proven science rather than trends.

Supporting Resource: Nicole Forsgren's original peer-reviewed papers on DevOps Research & Assessment (DORA)

Synthesising with Broader Industry Practice

The principles of Accelerate align with several other transformative bodies of work:

  • SAFe (Scaled Agile Framework): Alignment with metrics and CI/CD focus
  • Lean Startup by Eric Ries: Emphasis on fast feedback and learning loops
  • Working Backwards by Colin Bryar (Amazon): Customer-centric product thinking
  • The Goal by Eliyahu Goldratt: Emphasis on flow and systems thinking
  • SRE (Google): Emphasises error budgets and reliability engineering, dovetailing with lead time and failure recovery

ℹ️ Action:

Integrate Accelerate’s capabilities with your agile and product