Analyzing Business Data with Excel, by Gerald Knight, ISBN: 0­596­10073­6

Reviewed by: Robert Boardman, June 2006
Published by: O'Reilly
Requires: N/A
MSRP: US$39.99, CAN$55.99

This has been an interesting, important and somewhat difficult book to review. It's not a textbook or guide or manual about learning how to use Excel in business and there is little material to help business users make the transition from competence with Excel to data analysis and situational modeling with Excel. As the author writes in the Preface, "This book is written for experienced Excel users. It doesn't spend much time on basics and assumes the reader already knows how Excel works. Most of the chapters in this book start with a business problem or question, and then show how Excel can be used in that situation.” And there are quite a number of problem/solution sets. The book is interesting and important because author Gerald Knight uses Excel for tasks that are not often explained in books.

Knight starts with a quick introduction to array formulas, cell addressing and statistical functions. Then following chapter is about using pivot tables, an essential element in data analysis with Excel. With these tools and others he then examines specific problems and suggested solutions for workload forecasting, modeling complex situations, statistical process control, monitoring complex situations, work process in queues, and optimization using Excel.


Knight discusses the difficulties presented when importing data into Excel and some techniques to use when presenting data and solutions to others. Along the way he develops some complex and useful Excel files and tools, including a couple of interesting macros. These completed tools are provided as files that can be downloaded from the publisher's site.

Knight uses real data—lots of real data—to generate useful solutions to real problems. For example, chapter 3 is entitled, “Workload Forecasting”. Knight uses data which he says is from an actual call center. The data is the workload for 116 days—five months of calls in 1998. The goal of the project is, “to predict a periodic workload that could have an overall trend, and is subject to short-term ups and downs. The accuracy of predictions will be measured and used to set a prediction range with a known probability.” (p.43) Since this is the first chapter where he develops an application, he describes in some detail the steps and includes clear instructions about what to type in which cells along with the complete macro. The downloadable files are not locked or protected, so readers can use the book with the file open in Excel and follow along. Knight includes several suggestions at the end of the chapter for improvements, additions and changes that might make this application more useful in a particular situation.

Knight does not go into the cell-by-cell detail in the rest of the chapters as he does in chapter 3. Since all the applications he develops are completely open, there is very little need for that. The macros he writes are fully commented and do not need to be accompanied by additional instructions.

As I read and worked with the book and its files, I kept thinking there was something missing. Something about the book and files bothered me. I think the issue for me is that I want to be involved when I learn and I was not, at least not initially. Part of the reason for the lack of involvement is that the applications are presented complete and useable. I would have liked to see some of the development steps along the way either illustrated in the book or in the files. Knight probably feels advanced Excel users do not need this much support. Some of the connections between the parts of each application were less than clear to me in part because of the author's assumptions. Not only does he assume readers have above average knowledge of Excel, he also assumes readers have a solid foundation in statistics and in statistical analysis of business problems. However, by the third or fourth time through the same chapter these issues started to disappear. More and more I was pulled into the examples and tempted to explore variations.

Author Gerald Knight writes using vocabulary and assumptions that not all business users will share. But the book is for those involved with forecasting and statistics, for those with piles of data to analyze and who are desperate for tools to use to take control of this information. Knight's book provides some very powerful and very useful applications and methods which will be of great benefit to the appropriate audience.

KSN Product Rating:





© Copyright 2000-2006 All rights reserved. legal notice
home | previous reviews | hot news | about us | search | store | subscribe


Forums Search Home Previous Reviews About Us Store Subscribe