Friday, October 14, 2016
Review of DATA DRIVEN, by Jenny Dearborn
Douglas Winslow Cooper, Ph.D.
"The Holy Grail of data analytics, at least with respect to sales, is to increase the effectiveness of the individual. When individuals are more productive, motivated, engaged, and happy, the organization will be more successful. The real power comes when big data gets personal." Thus does Jenny Dearborn, Senior Vice President and Chief Learning Officer for software giant SAP, describe the value of what she covers here.
I requested a copy of this book from a friend who was involved in its preparation. I have long had an interest in statistical analysis, and my recent venture into entrepreneurship, through my small business, WriteYourBookWithMe.com, made me curious about what such analysis could do for marketing and sales.
When I previewed the book, by examining the ratings it got on Amazon reviews, I was a bit surprised to see not only a series of five-star ratings each with typically a couple of reader endorsements of the value of the enthusiastic review, but also one two-star rating, with nearly a dozen statements of the value that the readers had obtained from that rather negative review. This strong difference of opinion grabbed my attention, and I sat down to read the book, which took about four hours. I'll summarize what the author has done, and then I'll tell you why I think there were such divergent evaluations.
The book is a well written combination of story and analysis, handsomely presented by the publisher, John Wiley & Sons, Inc., with numerous graphs, charts, and tables and a comprehensive index at the end.
The story starts with our heroine, Pam Sharp, in her new position as Chief Sales Officer of the mythical Trajectory Systems, meeting with a subset of the relevant individuals she manages, to discuss how to overcome the deficiencies that have caused a disappointing sales year. Each department head has a story that clears that department from responsibility for the disappointing results.
Clearly, Pam has got to find a way to analyze what has gone wrong, to convince her subordinates of her analysis, and then to determine and implement the policies, strategies, and tactics needed to turn things around.
Author Jenny Dearborn provides non-technical explanations and fictional examples of four uses of data analytics:
1. Describing quantitatively what happened
2. Diagnosing what went wrong
3. Predicting what lies ahead
4. Prescribing what to do.
Describing can be done with familiar statistical and graphical techniques.
Diagnostic analytics involves trying to determine why something has happened, typically relying on methods of showing relationships between variables, such as outputs versus inputs. Very often correlations are highlighted, but only some of these reflect causation. Others are coincidental or are products of being influenced by a common third factor.
Predictive analytics answer the question, "What could happen?" Dearborn points out this analysis may include "statistics, modeling, machine learning, and data mining." For technical details, she directs the reader to books such as Siegel and Davenport’s Predictive Analytics. Here and elsewhere she does not require the reader to understand mathematical equations, but she does give appropriate references.
Prescriptive analytics involves using mathematical models to determine the optimal choice among various options, and can be as simple as using multiple linear regressions or as complex as the kind of machine learning that is sometimes used for email programs to distinguish spam from desired communication.
As the story progresses, Pam and her group overcome some organizational and personal obstacles to implement, and then demonstrate the value of, data analytics.
One of the early challenges is to obtain the appropriate data. Dearborn lists the following advice:
· "Cast a wide net."
· "Consider exclusions."
· "Be sensitive to the sensitivities (and politics)."
· "Marshall the right human resources."
· "Communicate your needs."
· "Be patient."
The team's mythical mathematician consultant, Henry, describes the various pitfalls that erroneous data or hasty generalization can create.
The rest of the book shows improvement in sales revenue and the use of these techniques to help the sales representatives upgrade their own performances. While not technical in nature, the description does explain in detail how to implement the results of data analysis to further the goals of the company, the departments within the company, and the individuals.
The book will not in fact help small businessmen such as myself, because we lack the data that make data mining and analysis worthwhile. Somewhat larger organizations might well benefit from hiring a consultant to set up such a system and to implement it. Larger organizations still might find this book useful to introduce members of the affected groups to the possibilities and methodology of analytics and a proposed data analytics system. Even major corporations may have use for this book, depending on their current state of awareness regarding data mining and analytics.
So, the objection that this book does not show how to do data analytics is valid, but this was not the goal of the author. Nor does it give a real-life demonstration of the value of these methods. Rather, the book will serve to help convince a certain class of readers of the value of such techniques and perhaps induce them to obtain the expertise to apply them to their own situation, to become “data driven.”