“Data is the lifeblood of business.” It is a popular saying, and in today’s world it is truer than ever. Electronically controlled and connected equipment streams data about our physical world and business activities at an ever-increasing speed. Communication takes place in an increasingly digital fashion.

IDC’s 2011 Digital Universe Study found that the volume of the world’s digital data amounted to 1.8 zettabytes, which would “fill up 57.5 billion 32GB tablet PCs.” The size of this data universe is estimated to more than double in two years, outpacing even Moore’s law. This staggering volume of data simultaneously presents opportunities and challenges.

Data can be examined and analyzed, which can provide the business greater insights, lead to better, faster decisions and uncover potentially hidden data patterns and relationships. However, data is only useful — and able to support the previously mentioned activities — if it is kept organized and clean. Therein lies one of the many challenges associated with collecting and leveraging data. 

Improving and preserving the quality of data can be a daunting challenge. For most organizations, it is an investment of a magnitude that requires buy-in and approval at the executive level. This article is neither a discussion of ROI for sound data management practices, nor it is a step-by-step manual for implementing a data quality program. The purpose of this discussion is to introduce a viewpoint for confronting data quality challenges, especially the ones that can and will surface during any business transformation effort. The following outlines a business case for how business transformations present a particular opportunity to embrace — instead of dread — data quality challenges and obtain the executive-level buy-in for change.

What kinds of business transformations might present these opportunities? When we consider the data deluge, it becomes apparent that most business transformations fall into this category. A few examples include implementing a new CRM process, redesigning a corporate portal, launching a new data warehousing initiative and, implementing a new ERP system to improve existing processes and replace a web of special purpose legacy applications.

There are three key concepts to help formulate an appropriate perspective for handling the data quality challenges associated with business transformation projects.

1.Incorporate an assessment of data quality risks and impacts in the planning process of a business transformation.

Organizations approach these programs and initiatives in different ways, but there are some common elements. These activities involve some level of planning and accounting for necessary resources, as well as a timeline to accomplish the desired goals. Often, risks are considered and mitigating strategies are devised. These risks can be directly related to the data, but more often they are not.  While there is an increased awareness of the risks related to data quality, planning is often based on anecdotal evidence and tribal knowledge when it comes to data. The lack of a risk assessment — based on systematic data profiling and business rule-based data quality analysis — often correlates with data quality-related emergencies, which can happen at the most inopportune time: late in the initiative.

Why do data quality problems tend to be so insidious and latent in the initiatives? Most are familiar with the natural progression of programs that follow (sometimes very loosely) the accepted principles of a software development life cycle. According to this approach, activities usually happen in a methodical sequence that follows major phases: requirements gathering, designing, building, testing, acceptance and release, leading to project completion. Anybody who has participated in this “ritual” has an appreciation for the length of time and the amount of effort involved before the project enters the testing phase. Often, this is the first time that the “rubber meets the road,” from the perspective of the data. Testing activities can bring a very rude awakening: The realization that the data does not conform to the business requirements. While the degree of nonconformance varies from program to program, it can be quite significant. 

2.Establish a common agreement on what quality data means to the organization and the business processes.

A frequently heard argument is “But my data is clean in the current system…” While this statement might be correct on some occasions, it is still better to analyze and assess your data in a structured, methodical fashion by conducting a data quality assessment. 

But for the sake of our discussion, let’s assume that the above statement is true and the data is of pristine quality. Before moving forward, it’s important to  discuss the definition of quality data.

Out of the many existing definitions, J.M. Juran’s description of quality — first appearing in the third edition of “Juran’s Quality Handbook” — might be the most revealing: “Data can be considered of high quality when it is fit for its intended use in operations, decision-making and planning.” What does this broad definition tell us? As mentioned above, business transformations usually involve data transformations, the key word being “transformation.” This means that during these programs and initiatives, there is a need to rethink – at least partially – the way existing data is generated and used. Sometimes, it may even require entirely new data attributes to be collected. Therefore, the purpose of the data and the associated quality requirements will change. This leads to the conclusion that the current data — regardless of how well it satisfies the current information needs — might not satisfy the transformed business needs and processes. 

Arriving at a detailed definition of “high quality data” can be a tedious exercise. Organizations often have internal debates over the basic definitions of data objects, let alone the associated quality requirements. For example: How do we define our customers? Are reselling channel partners considered customers? Is a “lead” considered a customer if an official quote is extended? Is a customer considered active if the last order was three years ago?

The answers to these specific questions drive not only many different business processes, but also the definition of quality data. And the answers to these questions might be very different depending on whether you ask the marketing department, the sales department or accounts receivable. Each business group may have their own set of business rules and their own specific set of quality definitions that enhance a broader, common definition of enterprise data. These business rules can peacefully coexist and data can be evaluated against them from various perspectives, depending on the specific purpose and its intended use. Of course, when business rules create a conflict that impacts the actual value of a data attribute (e.g., active versus inactive), that conflict needs to be resolved through consensus, which often involves practical workarounds (e.g., creating custom reports, custom attribute fields, etc.). 

3.Do not expect data quality issues to magically go away once the initiative is completed.

We have examined how data poses a considerable risk and reviewed some of the challenges associated with defining quality itself. Now you might ask, so, what is the specific opportunity that the main title of the article implies?

Risks and issues stemming from data can and should be mitigated the same way as any other risk factors during the course of the initiative. With proper assessment, planning and action, data can be appropriately addressed. And therein lies the opportunity: The steps that are taken to assess the quality of the data — and subsequently remediate data quality issues — need to be treated as foundational stepping stones to something greater than the success of the current initiative. They should become the stepping stone for an ongoing data quality program, which is critically and strategically important to the enterprise in preserving and maximizing the value of its data assets. 

The focus for this transition needs to be on the process, the technical platform and the skill sets of the people that participated in the data quality assessment. These factors will become crucial components of an ongoing data quality program. An ongoing data quality and assessment program should be standard operating procedure for business organizations, much like network maintenance, security, internal and external threat assessment, or equipment maintenance and repair, or monthly closing and reconciliation of financial records. 

The one-time data activities associated with delivering business transformation projects should not be viewed as a necessary evil, but should be re-examined and leveraged to launch an enterprise data management initiative focused on maximizing the vast potential value of the data. Organizations should implement a well-thought-out methodological approach to data, similar to other quality assessment and maintenance approaches (e.g., Six Sigma).

The following action items should be considered when business transformation initiatives are planned and executed:

  • Projects and initiatives should include an assessment and mitigation plan related to data quality risks. 
    • While an interview-based risk assessment should be a component, it should not be the sole means to assess risk. Subjective bias and long-held corporate beliefs (i.e., the “collective consciousness” of the organization) may mask important risks. 
    • Objectively measure the quality of data through data profiling analysis and a business rule-based data quality assessment process. Evaluate and re-evaluate risks and mitigation plans iteratively, as more objective statistics become available about the quality of data. 
  • A dedicated team and team lead should be incorporated whose primary focus is data.
    • Enterprise initiatives frequently do not devote dedicated resources to data quality. This can push the issue of data quality to the back burner when process design issues, often perceived as more pressing, start to overwhelm the program. 
  • If the enterprise does not have an existing data quality program, the charter of the data team should include post-program responsibilities. The goal of forming an ongoing data quality team needs to be clearly communicated upfront. This will foster the development of an "investor" mindset, with a long-term vision in focus, as opposed to the "check the box" mentality with which data efforts are often plagued. 
  • Copyright (c) 2013 Deloitte. All rights reserved. Published with permission.

    Disclaimer:
    This publication contains general information only and Deloitte is not, by means of this publication rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional advisor. Deloitte shall not be responsible for any loss sustained by any person who relies on this publication.

    Register or login for access to this item and much more

    All Information Management content is archived after seven days.

    Community members receive:
    • All recent and archived articles
    • Conference offers and updates
    • A full menu of enewsletter options
    • Web seminars, white papers, ebooks

    Don't have an account? Register for Free Unlimited Access