Understanding the Predictive Analytics Lifecycle

By Alberto Cordoba

A high-level, informal look at the different stages of the predictive analytics cycle

Understanding the Predictive Analytics Lifecycle covers each phase of the development of a predictive analytics initiative. Through the use of illuminating case studies across a range of industries that include banking, megaresorts, mobile operators, healthcare, manufacturing, and retail, the book successfully illustrates each phase of the predictive analytics cycle to create a playbook for future projects.

Predictive business analytics involves a wide variety of inputs that include individuals' skills, technologies, tools, and processes. To create a successful analytics program or project to gain forward-looking insight into making business decisions and actions, all of these factors must properly align. The book focuses on developing new insights and understanding business performance based on extensive use of data, statistical and quantitative analysis, explanatory and predictive modeling, and fact-based management as input for human decisions. The book includes:

• An overview of all relevant phases: design, prepare, explore, model, communicate, and measure
• Coverage of the stages of the predictive analytics cycle across different industries and countries
• A chapter dedicated to each of the phases of the development of a predictive initiative
• A comprehensive overview of the entire analytic process lifecycle

If you're an executive looking to understand the predictive analytics lifecycle, this is a must-read resource and reference guide.

• Language: English
• Publisher: Wiley
• Release date: Jul 30, 2014
• ISBN: 9781118938928

Book preview

Contents

Foreword

Preface

Acknowledgments

Chapter 1: Problem Identification and Definition

Importance of Clear Business Objectives

Office Politics

Note

Chapter 2: Design and Build

Managing Phase

Planning Phase

Delivery Phase

Notes

Chapter 3: Data Acquisition

Data: The Fuel for Analytics

A Data Scientist’s Job

Notes

Chapter 4: Exploration and Reporting

Visualization

Cloud Reporting

Chapter 5: Modeling

Churn Model

Risk Scoring Model

Notes

Chapter 6: Actionable Analytics

Digital Asset Management

Social Media

Chapter 7: Feedback

What the Different Software Components Should Do

Note

Conclusion

Appendix: Useful Questions

Bibliography

About the Author

Index

End User License Agreement

List of Illustrations

Figure 3.1 SAS/ACCESS® Software

Figure 4.1 SAS® Visual Analytics Dashboard

Figure 4.2 Model of Actionable Data Discovery

List of Tables

Table 6.1 SAS Project Activity Allocation

Wiley & SAS Business Series

The Wiley & SAS Business Series presents books that help senior-level managers with their critical management decisions.

Titles in the Wiley & SAS Business Series include:

Analytics in a Big Data World: The Essential Guide to Data Science and its Applications by Bart Baesens

Bank Fraud: Using Technology to Combat Losses by Revathi Subramanian

Big Data Analytics: Turning Big Data into Big Money by Frank Ohlhorst

Big Data, Big Innovation: Enabling Competitive Differentiation through Business Analytics by Evan Stubbs

Business Analytics for Customer Intelligence by Gert Laursen

Business Intelligence Applied: Implementing an Effective Information and Communications Technology Infrastructure by Michael Gendron

Business Intelligence and the Cloud: Strategic Implementation Guide by Michael S. Gendron

Business Transformation: A Roadmap for Maximizing Organizational Insights by Aiman Zeid

Connecting Organizational Silos: Taking Knowledge Flow Management to the Next Level with Social Media by Frank Leistner

Delivering Business Analytics: Practical Guidelines for Best Practice by Evan Stubbs

Demand-Driven Forecasting: A Structured Approach to Forecasting, Second Edition by Charles Chase

Demand-Driven Inventory Optimization and Replenishment: Creating a More Efficient Supply Chain by Robert A. Davis

Developing Human Capital: Using Analytics to Plan and Optimize Your Learning and Development Investments by Gene Pease, Barbara Beresford, and Lew Walker

The Executive’s Guide to Enterprise Social Media Strategy: How Social Networks Are Radically Transforming Your Business by David Thomas and Mike Barlow

Economic and Business Forecasting: Analyzing and Interpreting Econometric Results by John Silvia, Azhar Iqbal, Kaylyn Swankoski, Sarah Watt, and Sam Bullard

Foreign Currency Financial Reporting from Euros to Yen to Yuan: A Guide to Fundamental Concepts and Practical Applications by Robert Rowan

Harness Oil and Gas Big Data with Analytics: Optimize Exploration and Production with Data Driven Models by Keith Holdaway

Health Analytics: Gaining the Insights to Transform Health Care by Jason Burke

Heuristics in Analytics: A Practical Perspective of What Influences Our Analytical World by Carlos Andre Reis Pinheiro and Fiona McNeill

Human Capital Analytics: How to Harness the Potential of Your Organization’s Greatest Asset by Gene Pease, Boyce Byerly, and Jac Fitz-enz

Implement, Improve, and Expand Your Statewide Longitudinal Data System: Creating a Culture of Data in Education by Jamie McQuiggan and Armistead Sapp

Killer Analytics: Top 20 Metrics Missing from Your Balance Sheet by Mark Brown

Predictive Analytics for Human Resources by Jac Fitz-enz and John Mattox II

Predictive Business Analytics: Forward-Looking Capabilities to Improve Business Performance by Lawrence Maisel and Gary Cokins

Retail Analytics: The Secret Weapon by Emmett Cox

Social Network Analysis in Telecommunications by Carlos Andre Reis Pinheiro

Statistical Thinking: Improving Business Performance, Second Edition by Roger W. Hoerl and Ronald D. Snee

Taming the Big Data Tidal Wave: Finding Opportunities in Huge Data Streams with Advanced Analytics by Bill Franks

Too Big to Ignore: The Business Case for Big Data by Phil Simon

Using Big Data Analytics: Turning Big Data into Big Money by Jared Dean

The Value of Business Analytics: Identifying the Path to Profitability by Evan Stubbs

The Visual Organization: Data Visualization, Big Data, and the Quest for Better Decisions by Phil Simon

Win with Advanced Business Analytics: Creating Business Value from Your Data by Jean Paul Isson and Jesse Harriott

For more information on any of the above titles, please visit www.wiley.com.

Understanding the Predictive Analytics Life Cycle

Alberto Cordoba

Wiley Logo

Cover image: © iStock.com/oliopi

Cover design: Wiley

Copyright © 2014 by Alberto Cordoba. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone], or on the Web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, [phone], fax [phone], or online at http://www.wiley.com/go/permissions.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at [phone], outside the United States at [phone] or fax [phone].

Wiley publishes in a variety of print and electronic formats and by print-on-demand. Some material included with standard print versions of this book may not be included in e-books or in print-on-demand. If this book refers to media such as a CD or DVD that is not included in the version you purchased, you may download this material at http://booksupport.wiley.com. For more information about Wiley products, visit www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

ISBN 978-1-118-86710-5 (Hardcover)

ISBN 978-1-118-93893-5 (ePDF)

ISBN 978-1-118-93892-8 (ePub)

Dedicated to my loving family

Foreword

With this book, Al has made an astonishing contribution to the growing body of knowledge about business analytics. The book offers an unprecedented look into the gory details of life as a programmer/analyst, BI specialist, researcher, project manager, data scientist, and consultant. It offers examples of problem solving that could only have been applied by using the progressive power of information technologies that was mastered in the 1990s and 2000s. Have there been other books about this topic? Of course, but none has portrayed the human side of this global endeavor with so much enthusiasm and humor before. For the first time, Al personifies the characters that play important roles in the lifecycle of the generation and use of predictive analytics, showing their creative abilities in industries such as banking, megaresorts, mobile operators, healthcare, manufacturing, and retail. These personifications helps us all better understand and manage this big and complex process of deriving information from data in today’s increasingly sophisticated race to drive productivity and innovation. This understanding is essential to excel at providing an outstanding customer experience, manage customer churn, and perform data-intensive marketing campaigns.

This book has many novelistic aspects and is very conversational. This way the stories do make each section more personal and relatable.

The Note Files in particular will be quite helpful for readers as they present examples of real-life analytics projects. These files can be used as good starting seeds for projects.

Leigh Watts

Preface

People get excited about big data and business analytics. That might sound ridiculous, but I’m not kidding. Over the years, I have traveled everywhere from Brazil to Japan working on analytics, and in every country, I have found a particularly peculiar brand of analytic fanaticism. These analysts are hilarious and downright exciting!

I wrote this book to help business and IT professionals understand the predictive analytics lifecycle. A reader can get a sense for the entire predictive cycle and therefore avoid potential risks. Often folks who work in a particular area of the analytics cycle—for example, analysis—have little understanding of another area, such as integration. This situation sometimes creates confusions, poor communication, and delays. This book has seven chapters, which illustrate a complete cycle from idea definition to feedback.

In each chapter, I added notes at the end with examples. The first chapter concentrates on the initial stages of defining the problem at hand and how to get the business customer to quantify the overall business value of the project. The note file contains a sample nondisclosure agreement. Chapter 2, the design chapter, focuses on the planning process. It shows how to define various components of the scope, such as what organizations are affected, what to expect, and the type of information required. The note file has a sample project for a data warehouse performance management project. Chapter 3, the integration chapter, discusses the process of bringing data together to build a file ready for analysis. This chapter includes two notes: One is a data-quality sample project, and the other is a file with a description of Hadoop and how it works with SAS. Chapter 4, the reporting and visualization chapter, illustrates how reporting and visualization techniques are used to review and make sense of data. The chapter includes a note file with an example of an analytic project focused on guest loyalty for a cruise ship. Chapter 5, the analysis chapter, presents how to build a couple of analytical models: a churn model and a scorecard. The note file presents concepts on fraud, waste, and abuse analytics. Chapter 6, about actionable analytics, describes how to use results from the predictive modeling in campaigns. There are two note files. The first has a simple assessment to identify gaps in a CRM analytics platform. The second is a sample project of the construction of a predictive analytics framework for a mobile operator. Chapter 7, the feedback chapter, discusses the iterative nature of the predictive analytics cycle and the importance of including feedback in the development of new models. The conclusion chapter provides a high-level view of the entire analytics lifecycle. The appendix contains more than 1,000 questions that can be used to qualify predictive analytics projects or simply to break the ice with both IT and business professionals interested in applied analytics projects.

This book is a mix of technical knowledge and business analytics humor. The names and the actions of the companies and employees have been changed in the interest of making the stories funnier and the content more readable. This book is for anyone who wants to gain a better understanding of the development cycle of business analytics in an entertaining way. Follow professionals of the Information Age as they tackle big data in this fascinating collection of case studies in different industries from around the world based on my real-life experience.

Acknowledgments

Since I began working with business analytics in 1985, I have had the good fortune to work and learn from some of the best minds in the world of data. When I joined SAS in 1993, I began to see the excitement that companies experience when they realize that they have finally found a way to use internal data to better understand both their organization and their customers and better manage their own performance by developing key performance indicators. I am grateful for the many conversations with my SAS colleagues and customers over these many years.

These conversations gave me the chance to see predictive analytics in action in different industries and different geographies, and helped me appreciate what predictive analytics contributes to the enhancement of customer experiences worldwide and how it generates value for organizations.

I feel fortunate to have this opportunity to say thank you to all of the amazing people who have coached me, including Jim Goodnight, Clive Pearson, Jeff Babcock, Eric Yao, Herbert Kirk, Lee Richardson, David Fender, Alan Spielman, Steve Gammarino, Rajani Nelamangala, Phil Hyatt, Helen-Jean Talbott, Chuck Zebrowski, Barrett Joyner, Leigh Haddon, Andy Bagwell, Jose Carvalho, Mariana Clampett, Barrett Joyner, Andre Boisvert, Carmelina Collado, Monica Grandeze, Marcos Arancibia, Kimio Momose, Carol Forhan, Bill Marder, Tony Pepitone, and Jon Conklin. Many more colleagues have contributed to my professional development, and I am grateful to them.

I can’t forget to mention my eldest daughter, Sienna, for her willingness to work with me on the project and her flexibility and insight into the manuscript. I would also like to thank my three younger children for their undying love and support: Ines, Sofia, and Diego. Finally, I want to extend a very special, love-filled thank you to my beautiful wife, Clara Maria.

Al Cordoba

Chapter 1

Problem Identification and Definition

How executives focus resources and assess an organization’s readiness for meeting the challenges posed by new business realities

Recently I met with a pair of business executives at the Gaylord Convention Center near Washington, DC.

Two analysts glided their way toward me. I smiled and went in for handshakes, exclaiming Hello there! Their names were Zizi and Javier. Both worked for a big corporation right outside of the Beltway in Maryland. I quickly launched into a flurry of business jargon, briskly walking toward the coffee kiosk, mouth running at a hundred million miles per minute. The executives shuffled after me, saying We are very interested in finding out more about developing a modern analytical system.

I bought a soy latte with an extra espresso shot. As the caffeine kicked in, I started by asking, What is your firm’s level of analytical maturity?

Javier looked at me and said, Before we get started, do we have an NDA in place? A nondisclosure agreement is a document signed to protect both parties. (A sample agreement is presented at the end of this chapter.) We sure do, I answered. Great! So let’s continue.

Javier stammered, I-I don’t know. I believe that analysis is a portion of the transformation cycle from data to knowledge to wisdom. So, probably the analytical maturity of an enterprise would tell how well it can leverage analysis and close the information gap. I am not sure where I would say our company is exactly.

My eyes met his as I popped a huge sparkly smile. Everybody knows the four key levels of an analytical framework are. . . .

I waited for a response. Zizi replied, Infrastructure, functionality, organization, and business, and these levels can be translated into an information evolution model for analytical applications.1

Javier piped up, What is the importance of this?

I answered, Those organizations that try simply to define and implement an advanced analytical solution in one step may end up taking far too long to finish building it and reap its benefits.

Zizi lowered her glasses and continued my thought seamlessly. And then, most likely, the analytical solution delivered will not meet needs because requirements usually change after an initiative is initiated or because the technology has already changed. We’ve been through that before.

Exactly! I added, There is an overarching need to build flexibility into contemporary analytical systems. Particularly now that data are growing exponentially and we are faced with big data everywhere. I believe enterprises need to assess the overall maturity of their analytical initiative and aim to add value incrementally rather than use an all-at-once approach. This is very important with the big data challenges. Results and challenges differ depending on the level of analytical maturity. I think the assessment of needs for an analytic platform or workbench should include choosing an appropriate software architecture for analysis and reporting, a hardware environment, a big data integration approach, and, of course, a data model for their structured data, among other things.

They wondered, Is that enough to ascertain success?

I told it to them straight. Hey, it’s anybody’s guess, but it increases the probability of success significantly!

Results usually are measured in terms of effective usage of information technology (IT) investments and improved operational efficiency. Challenges primarily occur with IT infrastructure, culture, software technology, and functionality.

They looked at each other warily. I tried to reassure them a little bit. Improved results usually are associated initially with having one version of analysis-derived information, the so-called truth, which improves the management of multiple departments. Some of the organizational challenges begin to take more focus and skills from the project team. Good results are associated with improved and faster decision-making processes than the competition.

I decided not to mention the challenges that often occur at the business level, such as shifting business processes and methodologies to leverage new analytical capabilities for corporate performance management. Or changing business goals or objectives, based on insight gained. They were too apprehensive. Therefore, I wanted to stick with the most basic and positive aspects of reworking their business objectives.

I continued, As your consultant, I have to ask you: Where is your firm going? Is the gut feeling still driving decision making? A successful analytical initiative needs good strategic business objectives.

They winced at that statement. They knew I was right. Javier shook his head back and forth and sighed.

IMPORTANCE OF CLEAR BUSINESS OBJECTIVES

I patted them on the backs. Business objectives must drive analytical initiatives and investments. The success of an analytical initiative should be measured by how it affects strategic and operational business objectives—not how many rows of big structured and unstructured data can be loaded into a data framework in six hours or the complexity of a model developed. This is particularly true when we consider the vast amounts of data that most organizations have accumulated and that continue growing.

Obviously, the lack of clearly defined business objectives would make assessing the success or value impact of an analytical initiative impossible.

Do you think that you can use an analytical framework to align IT system initiatives with business objectives and make strategic choices? I asked the executives. It could be the best thing that ever happened to you.

I recommended that they conduct a business value evaluation prior to investing

Reviews for Understanding the Predictive Analytics Lifecycle

[debashis · Rating: 5 out of 5 stars]

Very good book with real time industry example. Great book for young data analytics