In his new book, “The Big Data Revolution”, two top executives from the United States Federal Trade Commission discuss their vision of the future of work. Not long ago, we heard from several business professionals that they felt their position was at risk due to an inability to adapt. According to these individuals, their companies had become too dependent on technology and too inflexible in their processes. These executives believe that big data will be the key to unlocking the productivity hidden in everyday activity.
The authors begin by describing the data-driven decision making. This is a process by which management takes the actions necessary to improve or alter a company’s internal processes or business strategy. According to the authors, data makes it possible for people at all levels to participate meaningfully in the design of work processes and strategies. This provides them with the opportunity to build robust, creative ways of getting things done. Big data also allows for a company to look at existing strategies and determine whether any modifications are needed or if a total overhaul is needed.
The next two chapters describe how big data-driven decision making is not just good for managers. Employees also can benefit from data-driven decision making. Employees need to know what is going on within their own company. They also need to be able to visualize how their thoughts and feelings about work and the way they get things done may be shared by others within their organization. This chapter discusses some of the ways in which employees may make use of big data: using it to communicate, collaborating, and problem solving, as well as sharing documentation and process execution.
The fourth chapter focuses on why measurement error pervades every part of the scientific community. It explains three reasons behind measurement error and how measurement errors can be prevented. Authors explain how the scientific process works and present an outline of scientific methods. They then describe the problems associated with measurement errors and how the potential causes of measurement errors can be remedied. The authors conclude their chapter by recommending five ways companies can capitalize on big data.
The fifth and final chapter focuses on why big data has so many implications for business leaders and managers. Chapter six describes the main reasons behind messy decision making and examines the impact of unanticipated correlations. The author presents various ways people might think about and make decisions based on messy correlations. After describing the different ways people might think and combine data, they go on to describe two models that explain the phenomena and discuss implications. The author concludes this chapter by describing the implications of the models for leadership and managerial decision making.
This book provides very interesting insights into how business leaders and managers might use big data to solve small data problems, but also makes it clear that the data and its corresponding interpretations are not as simple or black and white as one would like. The authors describe potential sources of biases and errors, discuss what types of measurements are appropriate in a messy context, and explain what the implications are for decision-making. However, I think the biggest takeaway point is that many of the proposed solutions are only possible if measurement errors can be effectively controlled and measures can be reliably collected. The authors do describe a way in which managers can leverage big data to address measurement errors, but there are limits to this approach. For instance, the proposed solution to address the problem of measurement error involves creating elaborate metrics and relying on these measurements to inform strategic decisions.
The book ends with a short description of the authors’ work and some current related topics. While this book is well written and draws on some interesting current ideas, the authors claim that much more work needs to be done on measurement and statistical inference, and they suggest developing measurement techniques based on supervised learning. While this sounds like an interesting area for future research, I am not sure that I entirely agree with them on this last point, and I believe that some of their proposed solutions are already being used in practice. However, I think they make a very compelling argument and will stimulate people to continue to think about big data and its implications.
In this book, the authors make the case that big data is really a revolution that will transform how we live. The authors claim that there is plenty of room for improvement in statistical methods and measurement, and they provide some good examples of problems that have been addressed by successful companies using big data. Overall, this is a really good primer on the historical development of statistical methods and their applications, and the authors definitely present a clear case for why big data has so much promise and so many challenges.