Our online Data Quality Bookstore has taken on another recent addition with the release of "Executing Data Quality Projects: Ten Steps to Quality Data and Trusted Information" by Danette McGilvray of Granite Falls.

In this post we explore the content and methodology presented in the book.

The first thing that struck me upon obtaining the book was the size of the physical dimensions - this is now officially the widest and tallest data quality book in my possession!

This is in stark contrast to several of the data quality books I have read recently that resemble texts from academic research papers - crammed with small fonts, endless chapters, poor notation and sparse diagrams.

“Ten Steps” has taken a different approach and this is a refreshing change.

Large text, plenty of diagrams, simple navigation, “bite-sized” sections – full marks for presentation. Someone has clearly put a lot of thought into how this book will be used and it has paid off, it was effortless to read and follow.

Structure of the Book

The book focuses on a data/information quality improvement methodology Danette has termed “The Ten Steps” process.

Danette describes this herself as “…explicit instructions for planning and executing information quality improvement projects with detailed examples, templates, techniques and advice”.

In addition to this core framework are additional data governance related sections covering “Framework for Information Quality”, “The Information Life Cycle” and the “Data Quality Improvement Cycle”.

There are now several books that cover data quality assessment and improvement so it is a positive move to see some of the wider tools and techniques of data governance also included in a tutorial style framework focused on data quality.

The “Ten Steps” Defined

The key process phases Danette covers in her Ten Steps methodology are:

• Define Business Need and Approach
• Analyze Information Environment
• Assess Data Quality
• Assess Business Impact
• Identify Root Causes
• Develop Improvement Plans
• Prevent Future Data Errors
• Correct Current Data Errors
• Implement Controls
• Communicate Actions and Results

Although these phases are nothing new to most practitioners, the quality and completeness of the examples, materials and instructions certainly are.

David Plotkin (Data Quality Manager, California State Automobile Association) sums this up perfectly: “In a subject that is long on talk and short on practical advice for implementation, [Ten Steps] is a refreshing exception”.

This is a book that urges the reader to get out there and take action, not pontificate on the finer points of data quality theory.

Flexible and pragmatic

One criticism often leveled at data quality methodologies is that they lack flexibility and often appear  too “cast-in-stone”.

Danette has rightly pointed out that her data quality methodology provides sufficient instructions, techniques, examples and templates to provide enough direction to help the practitioner determine their own options.

As ever in these projects, it is up to you the practitioner to decide what activities are appropriate given the situation your organisation or client is facing.

In particular, the section on how to structure a data quality project and select the different phases depending on the exact nature of your challenge clearly demonstrates the flexibility of her approach.

Useful templates and offline materials

Danette makes considerable use of templates and examples throughout the text and this works well to communicate the concepts presented.

Clear diagrams, well annotated, using real-life scenarios bring the book to life and clearly demonstrate that this is an approach forged from many years working at the “coal-face” of data quality improvement.

However, what I really found valuable was the fact that the templates and much of the framework information is also available on the Granite Falls website designed to supplement the book.

(Note: click here http://tensteps.gfalls.com/ and you will see all the materials available. In addition, many of Danette McGilvrays internet publications are linked in our DQ QuickLink tool so you can get a good feel for some of the content in the book by reading some of her materials - click here).

Including additional materials is commendable and demonstrates that Danette is focused on supporting practitioners. She has clearly thought hard about how people can use the book for maximum effect.

Tip: By combining the book and the template materials you can start building out a data quality and data governance framework.

Why not combine these materials with our tutorial on how to create an online data quality management repository (click here for details)?

Benefits of Each Phase Clearly Defined

One aspect I particularly liked about the book is that it is pitched at varying levels of experience.

Throughout the book Danette explains in simple terms the benefits of the concepts presented. These may be obvious to experienced practitioners but if you were putting together a presentation to business sponsors for example, you now have a ready made set of materials that could be used to explain the benefits of your approach in a language everyone can understand.

Summary

In my opinion, Ten Steps is one of the most effective publications available on data quality to date.

I use the word “effective” because one of the core messages I took from the book is one of delivering what you can, with the resources and finances at your disposal using a straightforward approach that can be easily communicated.

If you are new to data quality you will enjoy this book immensely. I certainly wish it had been around when I started out.

The simplicity of the methodology combined with the meticulous attention to detail, structured steps and clear examples should make it easy for a beginner to learn and implement the techniques.

So what about the more experienced practitioners?

The quality of the methodology is such that I fully expect many consultancies or in-house teams to adopt and present it as part of their core approach.

On certain topics such as data profiling for example, clearly an exhaustive set of tutorials cannot be presented so there is certainly an opportunity for practitioners to combine their own knowledge and experience to create an even more comprehensive methodology with the Ten Steps forming a strong backbone.

By adopting the lessons presented in "Ten Steps" I believe that all data quality professionals will find this publication beneficial in helping take their career or business to the next level.

What do you think? Do you agree or disagree? Have you found the book useful? Why not add your comments below.