Mary Beth Wakefield
Vice President and Executive Group Publisher
Vice President and Executive Publisher
Joseph B. Wikert
Graphics and Production Specialists
Stephanie D. Jumper
Quality Control Technicians
TECHBOOKS Production Services
Richard T. Evans
Ive known Joy Mundy and Warren Thornthwaite for a number of years . I wont speak for them, but Ive reached the age where I dont want to remember how long Ive known people!
My respect and admiration for Joy began from the day we met at WebTV. Microsoft had recently acquired WebTV, so the SQL Server BI team went to take a look at their data warehouse. We quickly learned how much we didnt know. I met Warren slightly later at WebTV, where he impressed us deeply as well. Both Joy and Warren were clearly fine practitioners in the art of data warehousing.
After several years, Joy became a principal member of the SQL Server BI Practices team. While he was working independently, Warrens ongoing feedback and commentary on our SQL Server 2005 product efforts frequently led to changes in our thinking or implementation. Both authors have affected the direction of our product. We deeply appreciate their contributions.
In time, Joy joined Warren at the Kimball Group. We miss her still. On the other hand, we have always enjoyed a great relationship with the Kimball Group. Ralph Kimballs pioneering work in data warehousing has fundamentally influenced the SQL Server BI toolset in numerous ways. There are many Kimball Group Toolkits on my teams bookshelves.
Now, let me discuss the book you have in your hands. I love this book. When you build software, the thing you hope for most is that customers will use it in some way to improve their business, as in our case, or maybe even their lives. Time will tell if we have met that bar. When authors write books about software, they hope that their work will play a role in the readers quests. I think this book will help you get the most out of our software, SQL Server 2005.
The first ten chapters address designing and building a DW/BI system. The rest of the book discusses how to secure, deploy, and operate the system. The book is filled with prescriptive guidance and helpful hints. While the chapters are mostly independent of one another, I believe you should read them in order. Joy and Warren will teach you the Business Dimensional Lifecycle methodology from the ground up. Reading out of order or skipping chapters wont hurt you; I just think you get more than the sum of the parts by building your skills in lifecycle order. For example, in Chapter 2 the authors discuss data-driven models versus requirements-driven models. In following this discussion, you will be well served to have read Chapter 1 on gathering business requirements. Chapter 2 is a very strong tutorial on dimensional modeling. If you want to delve deeper, I suggest you take a Kimball University course, starring one or both of your authors.
Chapters 5 and 6 are important. They cover ETL (Extraction, Transformation, and Loading), a key to the success of any data warehouse or business intelligence project. Various estimates indicate that ETL work constitutes at least three- quarters of the effort of a successful data warehouse. Our goal with SQL Server Integration Services is to lower the ETL effort required. We built a highly productive environment for building ETL packages, including interactive and visual debugging of data flows. But we can go only so far with software. Joy and Warren provide practical advice on top of what we provide in the bits. For example, they recommend building a source-system sandbox when you find that your source exploration is going too slowly because of system load or is in fact adding to the system load.
Chapter 10 addresses data mining. At Microsoft we believe deeply in the potential of data mining in BI applications. Weve gone a long way, we think, toward making data mining far more accessible than ever before. Joy and Warren dont rehash the material you already have in Books Online. Instead, they teach you how to approach a data mining problem. Using case studies, they teach you how to think about and experiment with data.
I could list highlights from every chapter. But I suspect youre ready to get going on your own project! Let this book guide you along the way. Try to take Joy and Warrens advice; they have been building data warehouses and teaching data warehouses, successfully, for quite a while. We built SQL Server 2005 to help you succeed in your business. I could not be happier than to have Joy and Warren, and this book, help you get there.
Bill Baker (firstname.lastname@example.org)
General Manager, SQL Server Business Intelligence, Microsoft