Data Science and Analytics Strategy: An Emergent Design Approach

Description

This book describes how to establish data science and analytics capabilities in organisations using Emergent Design, an evolutionary approach that increases the chances of successful outcomes while minimising upfront investment. Based on their experiences and those of a number of data leaders, the authors provide actionable advice on data technologies, processes, and governance structures so that readers can make choices that are appropriate to their organisational contexts and requirements. The book blends academic research on organisational change and data science processes with real-world stories from experienced data analytics leaders, focusing on the practical aspects of setting up a data capability. In addition to a detailed coverage of capability, culture, and technology choices, a unique feature of the book is its treatment of emerging issues such as data ethics and algorithmic fairness. Data Science and Analytics Strategy: An Emergent Design Approach has been written for professionals who are looking to build data science and analytics capabilities within their organisations as well as those who wish to expand their knowledge and advance their careers in the data space. Providing deep insights into the intersection between data science and business, this guide will help professionals understand how to help their organisations reap the benefits offered by data. Most importantly, readers will learn how to build a fit-for-purpose data science capability in a manner that avoids the most common pitfalls. Cover Endorsement Half Title Series Information Title Page Copyright Page Dedication Table of Contents Foreword Preface Acknowledgements Contributors 1 Introduction Data Science as a Sociotechnical Capability The Fallacy of Strategic Alignment A Tale of Two Databases The Wickedness of Building Data Capabilities The Notion of Emergent Design What to Expect From This Book The Structure of the Book Notes References 2 What Is Data Science? The Data Analytics Stack Data Ingestion Storage Access BI vs Analytics vs Data Science Are You Ready for Data Science? The Data Science Process Doing the Thing Machine Learning Problem Types Test Your Knowledge What About AI? Great Power, Narrow Focus Doing the Thing Right Doing the Right Thing In Closing Notes References 3 The Principles of Emergent Design The Origins of Emergent Design Is There a Better Way? Emergent Design, Evolution, and Learning Uncertainty and Ambiguity Guidelines for Emergent Design Be a Midwife Rather Than an Expert Use Conversations to Gain Commitment Understand and Address Concerns of Stakeholders Who Are Wary of the Change Frame the Current Situation as an Enabling Constraint Consider Long-Term and Hidden Consequences Create an Environment That Encourages Learning Beware of Platitudinous Goals Act So as to Increase Future Choices Putting Emergent Design to Work – An Illustrative Case Study Background The Route to Emergent Design A Pivotal Conversation First Steps Integrating the New Capability The “Pilot” Project Scaling Up The Official OK Lessons Learnt Summarising Notes References 4 Charting a Course Introduction Tackling the Corporate Immune System Finding Problems Demonstrating Value Additional Benefits of “Problem Finding” Powerful Questions Designing for the Future Notes References 5 Capability and Culture Introduction Data Talent Archetypes Database Administrators, Data Engineers, and Data Warehouse Architects Key Skills What They Do What They Don’t Do Business Intelligence (BI) Developers and Analysts Key Skills What They Do What They Don’t Do From Analyst to Data Scientist Key Skills What They Can Do for You What They Don’t Do Newer Data Roles Other Roles The Right Timing Building Capability Identifying and Growing Talent Designing Training Programmes Immersion and Other Approaches Ongoing Development; “Building a Culture” Communities of Practice Data Literacy Critical Thinking Problems, Hypotheses, and the Scientific Method Measuring Culture Some Principles for Developing a Data Culture Buying Talent Find the Right Mix of Skills Value Problem-Solving Assessing Broader Skills Get the Candidate to Work On a Real Data Problem Get Them to Do a Presentation Match Expectations Consider a Broader Talent Pool Conditions Over Causes Closing Remarks Notes References 6 Technical Choices Introduction Cloud and the Future (Is Now) Cost Savings Simplicity of Setup Security and Governance Backups The Main Players API Services Code vs User-Centricity Guiding Principles Reducing Complexity Blocking Innovation Interoperability Usability The Effort to Keep Things Running Transparency Working With Vendors Initial Discussions Making the Choice Taking Root…and Growing (The Wrong and Right Way) Concluding Remarks Notes References 7 Doing Data Science: From Planning to Production Introduction The Machine Learning Workflow The MLWF in Depth Problem Formulation and Context Understanding Data Engineering Model Development and Explainability Deployment, Monitoring, and Maintenance Dialling It Down Concluding Thoughts Notes References 8 Doing the Right Thing Does Data Speak for Itself? Might Do Now, Must Do Later? Responsible AI Data and AI Governance Data Governance and Data Management Getting Started With Data Governance How Do You Figure Out a Direction? Data Stewards and Councils What About AI Governance? Trust Explainability and Transparency Bias and Fairness Acting Ethically Find What Matters Build in Diversity Establish Meaningful Rituals Continual Awareness Leverage Existing Tools (To a Point!) In Closing Notes References 9 Coda The Principles of Emergent Design Redux Selling the Strategy; Choosing Your Adventure Example 1: As a Means to Enable Data-Supported Decision-Making Example 2: As a Means to Support the Organisation’s Strategy Quo Vadis? Note Index