Data Governance Ebook Final

A De nitive Guide
to Data Governance
eBook
Table of Contents
1. Introduction 2. What Is Data Governance Really? 1. What Is Data Governance? 2. Understanding Data Governance 3. Scoping the Data Governance Initiative 1. Four Steps to Scoping Effectively 2. Selling Data Quality to Senior Executives 4. Data Quality versus Data Governance 1. Are Data Quality and Data Governance the Same Thing? 2. Organizing for Enterprise-wide Data Quality 5. Leveraging Technology for Data Governance 6. For More Information...
Section 1
Introduction
Agility Killer or No Brainer?
Have you recently been involved with a major initiative such as MDM, BI, or CRM and data quality or data governance seemed to be just an afterthought? It becomes clear to you that data quality was never included in the project in the first place. How many times have you actually implemented data quality solutions at a project layer, only to ignore the upstream and downstream interdependencies of data quality? If you can relate to any of these scenarios, youre not alone. In fact the industry is full of such examples, so it should come as no surprise that companies end up with project failures that put them in a position in which they cannot trust their own data. They just dont take the time to do it right the first time, at the outset, so that all successive results will map closely to intended project goals and expected results. Furthermore, there are excessive overhead costs associated with managing large systems, environments, and initiatives such as MDM, where data quality and governance have been absent. If you dont get the data quality right at the outset of the initiative, and if you dont set down rules for governing data and information, then youre setting yourself up for failure. Take the time to use tools and services to get the data right, map out a strategy and approach, and make improved data quality the very first goal of any larger initiative.
Section 2
What Is Data Governance?

Data governance has three distinct pillars: 1. Administrative data governance 2. Technical data governance 3. Business data governance Administrative data governance focuses on things such as direction, prioritization, scope, structure, organizational alignment, business case, and funding. It also includes other administrative issues like policy, procedure, roles, and accountability. This parcel of data governance requires executive and business leader support and participation. It also requires a distinct set of business leadership skills to formulate and implement. Technical data governance relates to activities around technologies and systems that facilitate data management such as data models, metadata management, workflow management, data quality tools, CRM, and BI. As you might expect, this segment of data governance should be led by IT. Business data governance pertains to business-centric activities that influence data outcomes, such as business process, business rules, data standards, stewardship, reporting, metrics, and other associated activities. These governance activities should be business led. However, this does not mean there are three separate data governance programs. One is dependent on the other, with administrative governance serving as the overarching authority over the entire program. By working within these three pillars, organizations can better clarify their business requirements; develop better strategies and plans to solve their challenges; and ensure that the right skills, people, and groups are involved in the right activities at the right time and for the right reasons.
Understanding Data Governance

The quality, reliability, security, accessibility, and usability of a companys data dictate the effectiveness of the organization in driving new business opportunities, delivering state-of-the-art customer service, managing risk and compliance, generating meaningful decision support models, and reducing operating costs. In turn, the overall condition of data assets is directly dependent on the companys ability to align people, processes, lines of business, and technologies. It requires that all of these things be working together to produce the desired outcomes that make data fit for its business purpose. The only way to align the interdependencies of disparate people, processes, lines of business, and technologies is through a wellorchestrated data governance program. Read more about the business drivers and best practices around data governance.
Section 3
Scoping the Data Governance Initiative

Four Steps to Scoping Effectively
As companies continue to figure out data governance and attempt to understand how it will impact and add value to their business, there really is no single magic formula for data governance. Its not like one implementation suits all. Everyone has their own slant on what data governance is and what is should be. As you can imagine, these definitions vary across organizations, and even among businesses within the same industry. Because of this phenomenon, the actual scope of each data governance initiative is trending toward touching a wide range of applications, functions, or responsibilities. Companies are trying to cover all their bases. For some organizations its all about stewardship, for others it may be data quality, and for others it is data modeling or metadata management. In some cases, it covers a cross-section of these activities, each with a different depth and breadth of involvement. Furthermore, scope can vary based on project and data types. For example, many data governance programs are launched in conjunction with CRM, MDM, BI, or DW, and can focus entirely on B2C, B2B, consumer, product, or content data. Collectively these disparities are a common point of confusion for organizations looking to draw parameters around their data governance program. One way to scope a data governance program at a high level is to define four specific areas of focus for both the short and long term:
1. Projects Identify the project that the initiative will begin with, as well as possible subsequent projects 2. Data Identify the data and the data sources that will be involved (consumer, products, content, etc.) 3. Data Management Function Identify each data management function that will be a part of the program both initially and over time (data quality, metadata management, data modeling, stewardship, lineage, security, etc.) 4. Business Function Identify the business functions that will be addressed (marketing, sales, customer service, risk and compliance, mail and transport, order to cash, etc.) Drawing much more manageable parameters around these aspects of data governance can be a useful high-level exercise, especially if it reflects both initial implementation and ongoing growth. It will also help you explain the short- and long-term direction of the program to executives, business leaders, and associates. And, even more important, by splitting the initiative into smaller bite-sized chunks, the project will be perceived as easier to implement, will offer better results quicker, and will help you obtain buy-in from key executives earlier in the process. And with buy-in comes funding and resource allocation, which is key to achieving your goals.
Selling Data Quality to Senior Executives

With over 15 years of data quality experience, Nigel Turner of NHT Data Consultancy has talked to many data quality practitioners and others who are trying to persuade their organizations to recognize that data quality improvement is a worthwhile investment. Across these discussions one common theme emerges. Although the benefits of data quality improvement appear self-evident to people intimately involved, a constant frustration is that others in the organization do not recognize its importance. This common complaint is directed most frequently at senior executives within an organization. People involved in data quality improvement know
that getting senior-level sponsorship and support is critical, but are frustrated that their efforts all too frequently fall on deaf ears. Senior executives just dont get it. The paper Reach for the Top: Selling Data Quality to Senior Executives helps to overcome this problem, and turn senior executives into allies. It is based on the authors experience initiating and leading a major data quality improvement program across a global telecommunications company, and then helping other large organizations tackle their data quality issues. The paper suggests strategies and approaches to help gain access to, and influence, senior executives within the organization. It also provides useful tools and techniques that can be employed before, during, and after engagement with senior executives.
Section 4
Data Quality versus Data Governance

Are data quality and data governance the same thing?
While both essentially strive for the same outcome of optimizing data and information results for business purposes, these disciplines address that goal from very different angles. They are not the same, but complement each other for the good of your business. From a high level, many companies may view data quality as a subset or part of data governance. In other words, data quality focuses primarily at a data element layer where technologies, best practices, and standards are applied in order to understand, analyze, monitor, fix, and report data anomalies within the data itself. Data governance deals primarily with orchestrating the efforts of people, processes, technologies, and lines of business in order to optimize outcomes around enterprise data assets. This includes, among other things, the broader cross-functional oversight of standards, architecture, business processes, business integration, and risk and compliance. In other words, it includes anything that can impact the integrity, quality, and security of company information. While data quality is mission-critical for any organization, it is often relegated to a single project or data source within the company. But this approach could severely restrict a companys bottom line and its return on investment for four primary reasons:
Project- or silo-based data quality prevents the company from understanding and addressing upstream problems that impact data outcomes caused by people and processes and that cannot be fixed by technology alone. Project-based data quality often leads to independent or isolated
data standards, business rules, and data models that cannot be fully leveraged across the company, therefore duplicating operational costs for future projects. Project-based data quality often leads to the implementation of multiple technologies for the same purpose within the company, which raises the cost of software, training, and maintenance. Enterprise-capable technologies are utilized in a single instance when they could be leveraged across multiple projects and data sources.
Data governance exposes these opportunities and risks, and provides the cross-functional platform necessary for organizations to address them accordingly. In other words, data governance enables enterprise data quality by breaking down the barriers associated with project-based data quality. All of this further improves data outcomes, bottom-line performance, and return on investment.
Organizing for Enterprise-wide Data Quality

Data quality has matured. In its early days the focus of efforts to improve the quality of data in organizations was predominantly tactical. Usually a specific data quality (DQ) problem was identified and a project initiated and delivered to resolve or ameliorate it. This approach was characterized by a heavy emphasis on data cleanse, a one-off process where shortcomings were recognized and quantified and improvements made. Although many organizations reaped rewards from this approach, often the underlying causes of DQ problems were at best partially addressed and were sometimes not tackled at all. The end result was that data cleanse became a regular, reactive, routine activity, with some data sources cleansed again and again. All too often the DQ improvement achieved was not sustainable. Moreover, these tactical approaches failed to recognize a critical truth about DQ, that the places in the organization where the problems were most acutely felt were often not the places where the problems originated. For example, incomplete or inaccurate capture of new customer names
and addresses by sales may not have a major impact on sales, but would undermine the efficiency of post-sales activities including invoicing, delivery, and product or service support. These shortcomings have led many organizations to move away from this tactical focus. Instead they have recognized the pervasive nature of DQ problems, and how an entire organization needs to be mobilized to fix them. To achieve this they adopt more strategic approaches, where truly sustainable DQ improvement results from cross-organizational collaboration and interworking. To make this happen, enterprise-wide approaches are put in place, including organization-wide improvement programs, pan-organizational data governance, and Master Data Management (MDM). This approach is not just quantitatively distinct from tactical initiatives but involves a radically different way of tacking DQ. This paper discusses the experiences of initiating an enterprise-wide DQ program at British Telecommunications plc. (BT). This program ran for 10 years, until 2007, when data quality initiatives were absorbed into business-as-usual activities. When it ended, BTs program had delivered over 75 data quality improvement projects, ranging widely in scope and purpose and affecting every BT line of business. Overall more than 625 million of verified cost reductions and other benefits were realized. Today many of these initiatives are still in place and continue to help BT maintain its quality of data. In addition, the culture of the company has changed to one where data quality is recognized as an essential prerequisite of an efficient business. The white paper Organizing for Enterprise-wide Data Quality Improvement:
Highlights BTs experience of organizing for enterprise-wide data quality Lists the key organizational structures Outlines the roles of the potential main protagonists and Summarizes the main lessons learned
Section 5
Leveraging Technology for Data Governance

How to Leverage Technology for Data Governance
Technology is a valuable component of a data governance foundation: a data discovery tool provides insights into what problems exist in a measured, quantified, auditable way; data quality processes apply multistep data quality checks, validations, corrections, etc., and ensure consistency across multiple systems; dashboards display results for a larger audience; and a data quality platform enforces data standards across an enterprise. Data Discovery Tools: These tools enlighten organizations as to what data quality problems exist, based on the results evident within the data itself. They allow data governance to be approached in a measured manner. They monitor and formally measure the degree to which standards are being adhered, across multiple systems and over time. They are used to create tangible inputs (reports based on the data itself) for data governance meetings so that meetings can focus on prioritizing known problems instead of recording anecdotally reported problems. Data problems can then be prioritized so that issues with the widest-reaching impact and the most significant business value are addressed first, and issues with quantifiably less impact after. Automated Data Quality Processes: Automated data quality processes are most effectively managed using data quality tools. These processes can range from traditional back-end name and address cleansing routines to front-end name and address cleansing workflows enforced upon each transaction. They may enforce customized product data quality standards or function as real-time data validation and correction routines. These processes may be single or multistep processes that can be reused to provide consistency across the enterprise.
Communicating Results: Dashboards show the quantified results of processes across time, and bridge the communication gap that often exists between IT and senior management. A Data Quality Dashboard provides a graphical display that shows how process improvements perform over time as well as how they perform against goals. Scorecards, trend information, and business rule performance can all be displayed graphically to immediately show impact in a measured way. Data Quality Platform: Data quality processes should be built and measured using tools on an enterprise platform suitable for growth and change over time. They must allow a way to implement a consistent process in more than one environment, system, application, or operating system, and then to expand as business needs expand. They should provide user interfaces that are suitable for business users to create standards and rules, define new processes, and manage changes. Tools should be scalable from both a content and a performance perspective, meeting the demands of a volatile business world and the increasing volumes of data that businesses create and use in todays global environment.
For more information

Harte-Hanks Trillium Software www.trilliumsoftware.com Corporate Headquarters +1 (978) 436-8900 trilliumsoftware@trilliumsoftware.com
______________________________________________________________________________
______________________________________________________________________________
Understanding Data Governance

Trillium Software Data Governance
Data Governance means different things to different organizations. In some cases, it even varies by industry. For many companies, Data Governance is directly related to the scope of an existing project versus the role governance can or should play enterprise-wide. At Trillium Software, we approach Data Governance at a holistic, enterprise level, knowing that for most organizations, this is the end goal and not something that is achieved overnight.
Data Governance is a formal function within a Company tasked with aligning people, process, and technology to administer and manage the Organizations data and information assets in a way that supports the overall business goals and objectives of the Enterprise. It specifically focuses on developing and implementing strategies, programs, and standards that promote data quality, integrity, accuracy, usability, reliability, security, consistency, compliance, and accountability across the Company.
The Reason for Data Governance

The quality, reliability, security, accessibility, and usability of a companys data dictate the effectiveness of the organization in driving new business opportunities, delivering state-of-the-art customer service, managing risk and compliance, generating meaningful decision support models, and reducing operating costs. In turn, the overall condition of data assets is directly dependent on the companys ability to align people, processes, lines of business, and technologies. It requires that all of these things be working together to produce the desired outcomes that make data fit-for-business-purpose. The only way to align the interdependencies of disparate people, processes, lines of business, and technologies is through a well orchestrated Data Governance program.
Fitting Data Governance into an Organization

Just like other company assets, data assets require an oversight body to properly organize, administer, manage, and coordinate activities and functions. In the case of data, this responsibility is assigned to a group commonly referred to as Data Governance or the Data Governance Organization. The table below outlines common oversight bodies and the assets they typically manage. The major difference between Data Governance and that of other asset management functions is that Data Governance is seldom a stand-alone organization within the company structure. Rather, it is mostly a decentralized operation comprised of a collection of individuals across the business. By its nature this makes it more difficult for companies to develop and implement effective governance programs that carry the responsibility, authority, and accountability required to be successful.
Trillium Software ______________________________________________________________________________
Common Oversight Body/Organizations for Company Assets

Finance
Asset Managed
Processes/Activities Administered/Managed
Financial (Cash, AR, AP, etc.) Capital equipment, facilities, furniture, etc. People Data
Order to cash processes
Asset Management
Capital equipment, facilities and the security and maintenance around them Hiring, benefits, evaluations, comp plans, etc. Standards, business processes, data modeling, business rules, security, compliance, stewardship, etc.
Human Resources Data Governance
Table 1. Common Oversight Bodies and the Assets They Typically Manage
Leading Data Governance within an Organization

Data Governance leadership varies across companies and industries. The most successful programs emerging today are led by business teams and are based on a democratic versus an autocratic approach to leadership. While many companies are looking toward their Finance organization to lead and/or chair their governance function, the finance and banking industries commonly turn to the Risk Management arm of the organization to lead their initiatives. Trends in this area indicate that businesses are recognizing that data assets are the responsibility of the business and not simply a task or activity to be managed by Information Technology.
Scope of a Data Governance Program

The scope of data governance programs within organizations vary by company size, industry, business requirements, budget, projects, span of control, and other factors. Regardless of enterprise size, every data governance program should have two foundational aspects that define program scope. The first refers to the scope of administrative functions and the other to the domains or disciplines it will oversee. Administrative Functions of a Data Governance Organization Well-designed data governance programs ensure that the organization has administrative responsibilities and authorities over its span of influence/control and disciplines to include the following: Strategy and Planning Roles and Responsibilities Scope Prioritization Business Alignment Funding Resource Alignment Business Cases Conflict Resolution Reporting
______________________________________________________________________________
2008 Harte-Hanks Trillium Software. All rights reserved.
2 of 6

Data Governance Domains/Disciplines From an enterprise perspective, a data governance program should bring structure, strategy, and continuity to the following areas of the business in order to drive the desired outcomes around data. Data Standards Business Rules Business Process Data Architecture Technology Data Quality Stewardship Analytics, Metrics, and Reporting Acquisition, Merger, and Enrichment Risk and Compliance Security
Data Governance
(Common Domains/Disciplines) Data Standards Consistency across systems for customer data, product data, business data
Business Process
Best practices and policy around data entry, validation, access
Business Rules Data Architecture Technology
Data processing, transformation, formatting, cleansing, matching
Data modeling and integration
Data quality tools, MDM, CRM, ERP, etc.
Data Quality
Data quality framework, processes, methodologies, and applications Alignment of education and best practices around business value, data profiling, tool usage, exception handling, and reporting Metric development, processes, content, and tools for analytics & reporting Evaluation and processing of new data sources from acquisitions and mergers, 3rd party data enrichment solutions/services Regulatory compliance (Sarbanes-Oxley, Basel II), audit requirements, credit risk
Stewardship Analytics, Metrics, Reporting Acquisition, Merger, Enrichment Risk & Compliance
Security
Data access rights and privileges, data storage, back up procedures, data import/export policies
Table 2. Data Governance Domains/Disciplines
______________________________________________________________________________
3 of 6
Trillium Software ______________________________________________________________________________ Data Governance Drivers

Trends indicate that more companies are now exploring the benefits that result from data governance due to the emphasis placed on enterprise data management systems, such as Master Data Management (MDM) and other solutions that incorporate multiple systems and lines of business. As the industry matures, many businesses recognize that in order to perfect data quality outcomes, they must address people, process, technologies, and change management in a consistent manner across the organization. For a growing number of companies, this means the starting point for governance is no longer at a departmental level, but rather at a larger scale within projects with far more visibility. Companies are also adopting Data Governance as an internal control to support the management of risk and compliance as it relates to Sarbanes-Oxley (SOX), Basel II, and other regulatory and audit requirements.
Benefits of Enterprise Data Governance

Enterprise Data Governance makes it possible for companies to orchestrate cross-functional activities that influence data outcomes, which in turn, drive business effectiveness around decision support, campaigns, customer service, operations, compliance, and risk. Effective enterprise data governance practices are aligned with internal business cases and deliver several key benefits: Drive mission-critical standards around data, business rules, business process, data architecture, and technologies across the company Create transparency into the companys data and processes, which in turn reveal business opportunities and challenges for the company Serve as a forum for establishing the responsibility, authority, and accountability needed around data management Develop a central repository of information for understanding, documenting, and acting upon complex data relationships across the company Promote education, awareness, and mindshare across teams and lines of business Provide a platform from which to scope, plan, and set priorities around data management and data quality initiatives Act as the administrative body for managing data acquisition and mergers Foster education, continuity, and scalability to a stewardship program Secure funding and resources for data related projects and programs Demonstrate and promote the business value of data assets
______________________________________________________________________________
4 of 6
Trillium Software ______________________________________________________________________________ Who Should Consider Data Governance?

Any organization considering or presently implementing a data quality or data management program spanning multiple business units and technologies should investigate data governance strategies to leverage time and resource investments. Such endeavors include: MDM, CRM, DW, CDH, ERP, BI Enterprise/large scale data quality initiatives Any data initiative intended to provide a single customer view (B-to-C or B-to-B space) Data initiatives focused on reducing financial risk and/or achieving regulatory compliance Projects intended to bring information together from several disparate data sources
Mission-Critical Best Practices for Data Governance

Key best practices apply to Data Governance at both a strategic and a tactical level. A handful of missioncritical best practices aligned with the strategic layer are certain to steer the success of any data governance program. Leverage Third Party Assistance An experienced 3rd party consulting group will help navigate the cultural and political barriers that are seldom overcome through traditional methods with internal resources. In addition a 3rd party will accelerate putting many of the following best practices in motion. Executive Support and Participation Mandate executive support and active participation to maintain interest, awareness, and visibility. Business Value Demonstrate and accentuate business value at every program phase. Decentralized Model Leverage existing company resources and reporting structures to build out the data governance program. Adding a new department with several new resources is not realistic for most organizations. Democratic vs. Autocratic Structure Architect a Data Governance structure where responsibility is shared across many people or small groups and rather than one individual. Center of Excellence (CoE) Incorporate a Center of Excellence within the framework of the broader Data Governance organization. Responsibility, Authority, Accountability Build clear responsibility, authority, and accountability into the Data Governance structure and individual roles. Communication Plan Adopt a well-defined and fluid communication plan that addresses executives, the Data Governance organization, and company associates. Education, Awareness, Mindshare Conduct interactive education sessions, seminars, and/or workshops to ensure that everyone involved understands all aspects of the program, their role, and the value they contribute. Data Governance Charter Create and actively maintain a Data Governance Charter that defines all aspects of the program and can be used as a tool to bring new people up to speed quickly as they rotate into the program.
______________________________________________________________________________
5 of 6

Momentum Assume that the program will lose momentum at certain junctures so plan ahead to avoid falling into a void where people loose interest. Small Administrative Footprint Develop a governance organization that involves the right people at the right time for the right reasons and in the right numbers to keep meetings and time commitments at a minimum and to achieve actionable results.
To learn more about: Understanding Data Governance Trillium Softwares Data Governance - Philosophy - Structural concepts - 9 step methodology - Scope and deliverables
Please contact: Trillium Software Sales: Jim Orr, Director Enterprise Strategy: Or visit www.trilliumsoftware.com
(978) 436-8900r (719) 481-0565
______________________________________________________________________________
6 of 6
Reach for the Top

Selling Data Quality to Senior Executives
Harte-Hanks Trillium Software www.trilliumsoftware.com Corporate Headquarters +1 (978) 436-8900 trilliumsoftware@trilliumsoftware.com EMEA +44(0)118 940 7600 Central Europe +49(0)7031 714756
TRILLIUM SOFTWARE
Introduction
I have worked in data quality (DQ) for over 15 years. In that time I have talked to many DQ practitioners and others who are trying to persuade their organizations to recognize that DQ improvement is worth investing in. In these discussions one common theme often emerges. Although the benefits of DQ improvement appear self-evident to people intimately involved, a constant frustration is that others in the organization do not recognize its importance. This common complaint is directed most frequently at senior executives within an organization. People involved in data quality improvement know that getting senior level sponsorship and support is critical but are frustrated that their efforts all too frequently fall on deaf ears. Senior executives just dont get it. This paper tries to help those working in DQ to overcome this problem, and turn senior executives into allies and not blockers of DQ improvement. It is based on the authors experience of initiating and leading a major DQ improvement program across a global telecommunications company, and helping other large organizations tackle their DQ issues. It will suggest strategies and approaches to help DQ professionals gain access to, and influence over, senior executives within their organization. It will also suggest some useful tools and techniques that can be employed before, during and after engagement with senior executives. The paper will also make reference to some of the main practical lessons learned by the author in running enterprise wide data quality initiatives.
Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com
TRILLIUM SOFTWARE
The Problem
An old South African folk tale relates the story of the Lost Message. This refers to the life of ants. Ants have many enemies man, birds, anteaters, centipedes so some of the ants decided they needed to improve their chances of survival by cooperating to repel their enemies. They held a council to discuss. But the council was a babel of discord. They talked together for a long time but nothing came of it. Eventually the council broke up without agreement and different groups of ants resolved to go their own way. As a result each had their own self-appointed tasks; workers, soldiers and so on. This division of labor preserved harmony amongst the ants but did nothing to protect them from their enemies. One day a king ant emerged. He asked the groups why they had not embraced the secret of unity and the benefits of working together. Each group gave him a different explanation based on their own roles in the kingdom. The king became very confused. Today ants are still to find the secret of unity and are still at the mercy of their enemies. Does this feel familiar? Our organizations have become increasingly diversified and more specialized. Each group of specialists within the organization has developed its own language, concepts and ways of working. Its become increasingly difficult for any individual to gain a picture of how it all fits together. In this context, pity the senior executive whose job it is to bring unity and harmony to these organizations. Every day he or she will be lobbied by one or more of these groups sales, finance, human relations, IT who will seek support for their investment cases and problems. He or she will be asked to make priority calls, favor one initiative over another, and act in the best interests of the organization as a whole. Its hardly surprising therefore that grabbing the attention and support of senior executives is a challenge. And its even more difficult for data quality professionals. DQ is a relatively recent discipline, and its unlikely that senior executives have been exposed to it in their education or experience, unlike sales, finance, human relations and so on. So the sell is even harder. DQ professionals must devise strategies to surmount this.
TRILLIUM SOFTWARE
Why should senior executives care about data quality?

First of all, why should senior executives care about DQ? The good news for DQ is that there is a plethora of evidence to demonstrate they should. Here are some examples:
In the current recessionary and stringent cost cutting environment facing both private &
public organizations, senior executives need to focus on minimising cost. Several surveys have demonstrated that poor quality of information costs the average organization between 10-20% of their revenue / turnover.
In the UK a 2008 Cap Gemini report Information Opportunity Report: Harnessing Information
to Enhance Business Performance found poor management of data and information cost the UK economy 67 billion a year. 46 billion of this was cost incurred in the private sector, 21 billion in the public sector.
In a 2009 Information Difference survey of global organizations with turnovers of $1 billion or

more, a third rated DQ in their organization as poor at best. Only 4% rated it as excellent.
In a 2009 Gartner survey of global CFOs 75% cited information as a barrier to achieving
business goals.
And these failures are occurring in the context of a data tsunami where data volumes are
exploding. By 2015 the average organization will hold 700 times the volume of data it held in 2000. By 2020 the average organization will hold 7,000 times the data held in 2000. Overall, coherent and integrated attempts to improve DQ in organizations could reap massive rewards, benefitting customers, employees and stockholders. Senior executives should therefore care very much about DQ.
TRILLIUM SOFTWARE
Why are senior executives important to data quality?

DQ improvement is a business transformation challenge. Poor DQ is usually caused by business process, people and technology issues coming together to create DQ problems. Resolving them therefore requires changes to business processes, to the way people behave and what they do, and to the technology that underpins business operations. In many cases resolving DQ problems requires strong collaboration and cooperation across the organization as the places where the pain is felt is often not where the problems originate. To tackle DQ across an organization consequently requires a transformation across the business, transcending organizational boundaries. Senior executives are critical in making this happen. They have a number of essential roles to play:
They will need to endorse the business case for DQ improvement and to sanction the
business investment to make it happen. Chances of gaining this approval will clearly be much greater if the senior executives involved understand the impact of poor DQ and the potential benefits of improvement.
As a business transformation challenge DQ improvement requires champions and mentors

at the most senior levels of the company. They can ensure DQ initiatives align with key current and future business strategies and goals and advise DQ teams on how to make the maximum impact across the business.
Senior executives are usually highly effective change agents. They are strong influencers
and negotiators who can open doors and break down barriers across the organization.
Through having direct authority over the areas of the business they lead senior executives
can issue direct instructions to their people to act in support of DQ improvement efforts. They can also mandate changes to business processes where required. To make DQ improvement happen the stick is often a useful adjunct to the carrot.
TRILLIUM SOFTWARE
Senior executives are often held up as role models. By themselves showing public support
for DQ improvement they can inspire others to do the same. A common barrier to DQ in many organizations is that its culture may be one where admitting to having DQ problems is seen as an admission of failure. By setting the example, senior executives can help remove this obstacle.
So why is it so hard to engage senior executives?

Weve already established that senior executives need better DQ, and that DQ needs senior executives. So why is this virtuous circle, clearly a win-win scenario, often so difficult to create and sustain? Lets look at some of the factors that get in the way:
The most obvious is that senior executives are busy people. In the daily course of their
activities they meet many people, are lobbied by many individuals and groups, and asked to make decisions about a wide range of topics. In their eyes DQ is another of many issues they need to face. Unless a case for action can be presented in a succinct, concise way there is little chance it will hit home.
As was touched on earlier DQ is a relatively recent discipline. It is unlikely that many senior
executives have encountered DQ as a business issue in their education or experience. To compound this, explaining the impact of poor DQ on a business can be complex, especially in larger organizations. The fact that it can and usually does have effects across and beyond an enterprise can add to this sense of complexity. When faced with priority calls between complex propositions and simple decisions, many senior executives will favor the latter. The implication of this for DQ professionals is that they must develop influencing strategies which simplify the issues, and relate their work to the issues senior executives face on a day to day basis. If they cant do this, their chances of successful lobbying are massively diminished.
TRILLIUM SOFTWARE
Finally DQ professionals must shoulder much of the blame. Many DQ specialists in

organizations have become, like the specialist ants, a subculture within their organization with their own mind sets, concepts and language. Too often, when given the opportunity to influence senior executives, they fail to translate what they are doing into language that a senior executive would relate to. Phrases such as metadata management, information architecture, data modelling, data governance, master data management et al are rooted in IT and are likely to elicit a response of blank incomprehension in many senior executives whose everyday business language centres on profit, brand value, process, revenues, costs, bottom line, customers, & shareholder value. To bridge this gap DQ professionals must therefore translate their everyday working concepts into language which can influence a senior executive, and which can elicit a positive response. To summarize, the problem faced by DQ professionals trying to gain traction and support at senior levels in their organization is qualitatively no different from that faced by finance, sales or HR professionals. They must be able to sell their case to senior executives in a way that will generate the support they need. But its a tougher sell, so DQ people need to work harder at it. Well now go on to propose strategies and techniques to help do this.
TRILLIUM SOFTWARE
The Solution How to Sell Data Quality to Senior Executives

For effective & enduring engagement with senior executives DQ professionals need to ensure they can tick all the following boxes. Lets call these The Six Steps to the Top:
STEP 1 - ARTICULATE THE PROBLEM

Understand the causes of DQ problems and be able to explain them in a way that will have resonance with senior executives.
STEP 2 CREATE THE VISION

Set out a clear vision of what DQ improvement activities are trying to achieve and have prepared 2, 10 and 30 minute pitches to communicate that vision.
STEP 3 DEMONSTRATE DELIVERY

Have a track record of delivering at least one, and preferably more than one, successful DQ improvement project which has impacted the businesss bottom line before engaging with senior executives.
STEP 4 - PREPARE THE GROUND

Are fully briefed and prepared for initial contacts with senior executive stakeholders.
STEP 5 - INFLUENCE THE OUTCOME

Can practice effective negotiating and influencing skills & behaviors when meeting senior executives.
STEP 6 - SUSTAIN INVOLVEMENT

Develop and implement organizational structures for DQ improvement where senior executives can play an active long term role. Well examine each of these steps in more detail:
TRILLIUM SOFTWARE
SIX STEPS TO THE TOP: Selling Data Quality to Senior Executvies

STEP 6 STEP 5 STEP 4 STEP 3 STEP 2 STEP 1
Sustain Involvement
Influence the Outcome
Prepare the Ground
Demonstrate Delivery
Create the Vision
Articulate the Problem
Step 1 Articulate the Problem

DQ specialists and others trying to gain traction for DQ improvement initiatives will readily recognize DQ issues. They also have practical experience of identifying, analyzing and articulating why poor DQ is happening. However many DQ people trying to drive change have their roots in IT; in many organizations DQ improvement still remains primarily driven by IT. This can create a particular problem when trying to articulate DQ issues in a way that senior executives understand. In many organizations IT continues to be seen as culpable for DQ problems. Many senior executives (apart from the CIO!) will therefore feel this is not their problem and will assume IT should and will sort it out. This of course is an erroneous perception of the causes of poor DQ. DQ is first and foremost a multi-faceted, inherently complex set of interrelated problems. This is illustrated on the next page.
TRILLIUM SOFTWARE
DATA QUALITY SYSTEMIC FAILURE
Human error No data accountability Poor training Internal politics Denial
DATA QUALITY
Poor process design Process failures Flawed goal setting No agreed data standards
PE OP LE
TECHNOLOGY
Data capture & U/D failures Multiple data silos Interface errors
S ES OC PR
This diagram demonstrates that:
The great majority of DQ problems found in organizations are usually the result of complex,
interconnected failures in peoples behavior, process design & execution, and underpinning technology.
Individual shortcomings in people, process and technology can themselves create DQ

problems but these are exacerbated by the fact that they often impact on each other, e.g. an incorrect input of an inventory item by a clerk will generate inaccurate information in the IT system which holds inventory information. This in turn can cause an automated process to fail, requiring manual intervention and incurring failure costs.
DQ people need to gain a good understanding of this holistic nature of DQ problems and
explain them in a way that senior executives will relate to.
10
TRILLIUM SOFTWARE
Success is explaining DQ in this way will convince senior executives that business leadership
of DQ is essential and so gain their endorsement and sponsorship. Being able to provide specific examples of failure is also obviously of great importance. Most important of all, DQ people need to understand business drivers and the impact on the business of the failures, both at a macro and micro level. To summarize, ensure that you understand and can explain in laymans terms the nature of DQ problems as interconnected failures impacting both business and IT, and caused in most instances by business and IT shortcomings.
Step 2 Create the Vision

As well as being able to explain both enterprise wide and small scale DQ issues and their impact in a way that will have resonance with senior executives, it is equally important to be able to paint a clear picture of what your DQ work is trying to achieve for the benefit of the business or organization you are operating in. To do this you need to have a clear, concise, communicable vision of what you are trying to achieve. This applies equally to a small, contained DQ improvement project right up to departmental or organization wide DQ improvement programs. This should consist of a short document or presentation which lays out:
Why DQ is important to your organization, or the part of the organization the initiative is
aimed at
The goals and objectives of your initiative A summary of the current situation with regards to DQ and its impact on the business,
expressed in business language
The anticipated benefits of your initiative, wherever possible in financial terms How you are going to deliver the benefits approaches, methods & tools
11
TRILLIUM SOFTWARE
Two key points to stress are:
To optimize the chances of gaining senior executive approval the document should relate the
initiative to key business goals, and/or the major strategic thrusts of the organization. This will ensure that it makes the DQ initiative relevant to senior executives.
All members of the DQ team should contribute to this visioning document and be able to
communicate it to all potential stakeholders and interested parties. The document should then form the basis of 2, 10 and 30 minute pitches that every member of the core DQ team should be able to deliver wherever and whenever the opportunity arises. The 2 minute pitch is particularly useful in planned or chance encounters with key senior executives where it gives the DQ team every opportunity to sell the merits of their initiative / project. Some DQ people have supplemented this with the production of flyers which can be handed out to senior people and all others who express interest or curiosity.
The vision should include a no action option. It is important to spell out the implications of
letting the status quo continue. What impact will this have on the current operational goals of your organization? What are the potential consequences for the future wellbeing and aspirations of the organization?
Step 3 Demonstrate Delivery

One of the most common mistakes made by DQ specialists trying to gain senior executive sponsorship and backing for their DQ proposal or project is to seek to elicit that support prematurely. At first sight this might appear to contradict received wisdom in DQ which states that senior executive support should be sought early. However it is borne out from the experience of many DQ specialists who have been down this route.
12
TRILLIUM SOFTWARE
Instead time your initial approaches to senior executives when you have hard, ideally quantifiable evidence to hand. This will give you the confidence to know that the vision you have laid out for your project or initiative can be substantiated and defended, and will also give senior executives more confidence in you. Moreover senior executives are confronted with people who bring them problems and are looking for solutions every day of their working lives. If you prepare the ground, you can bring them solutions, not more unwelcome problems. There are a number of things you can do to demonstrate the feasibility of your aims, methods and tools. These include:
Do data profiling of the key data of the problem area as early as possible. Obtaining and
applying data profiling tools are an essential prerequisite of success. These will enable you to scope and quantify the scale of the data shortcomings. Once this is done find out the impact of these problems on the key business areas by talking with people who depend on that data. In turn use this knowledge to assess the impact of these problems in business terms. The findings can help to refine the Step 2 Vision. It also provides you with the hard evidence to support your case to senior executives.
If your vision relates to a single DQ project, design and implement a trial or pilot to address
a smaller part of the problem area your intended project is trying to address. This will also enable you to test out your approaches and tools. For example if your aim is to enhance inventory data in a manufacturing company, profile data for one product, design a solution to improve the data, and assess the benefits of doing so. This proof of concept can then be used to refine the business case for wider action and provide solid evidence to underpin your case.
13
TRILLIUM SOFTWARE
Another useful precursor of selling DQ projects or initiatives to senior executives is to

use pilots or trials to establish relationships with key lower level stakeholders within their organizations. This can include both business and IT stakeholders. There are many benefits of doing this including:
These lower level stakeholders will need to be active participants in your initiative to
deliver the changes required to achieve your goals, so getting them involved early prepares the ground for later full scale delivery.
Lower level stakeholders can help you to quantify the impact of data quality problems
in business terms and so provide further credibility for your estimates.
They can open doors in their organizations to help you gain access to senior
executive stakeholders and advise you on the approaches to those stakeholders that are most likely to win approval see also Step 4 below.
Potentially they can accompany you in your subsequent meetings and contacts. You
are much more likely to gain the outcome you want with senior executives if their own people are alongside you demonstrating support for your proposal from within the senior executives own organization.
In summary a track record of delivery is in most cases a prerequisite of success with senior managers. In a larger DQ initiative the approach outlined above can be taken in early projects. In later encounters with senior executives a track record of delivery already established in other parts of the organization will give you the traction you need. Nevertheless the relevance of deliveries in other business areas to the particular senior stakeholder you are addressing must be demonstrated.
14
TRILLIUM SOFTWARE
Step 4 Prepare the Ground

Having laid out a clear vision of what you are trying to achieve and gathered hard evidence to substantiate it, you are now ready to take your case for DQ improvement to senior executives. When meeting senior executives it is far better for initial encounters to be on a 1-1 basis. This has the benefit of enabling you to tailor your approach to their individual needs. It also allows you to explore and discuss DQ issues that the senior executive may not want exposed in front of his / her peers. Furthermore you can take their potential objections to what you are trying to do on board and address them before moving on to other stakeholders. Crucially it de-risks this critical stage. Presenting your proposals to an organizational or departmental Board meeting with all key senior executives present could derail your initiative if you do not win favor and have allies on the Board to support you. Otherwise Groupthink or the herd instinct could dominate and one executives objections or reservations could unduly influence others. When meeting with senior executives for the first time it is always better to do so face to face whenever possible. This will help you to establish a personal relationship with the senior executive, important for building long term support. It will also enable you to present substantiating material prepared beforehand. If face to face is not feasible prepare some supporting material and either send it to the senior executive in advance of a call or present on a web conference. Whatever the method adopted, there are some pre-meeting activities that if carried out will increase your chances of getting the supportive outcome you desire. These include:
Do some preparatory stakeholder analysis. Stakeholder analysis is defined by Wikipedia as

the process of identifying the individuals or groups that are likely to affect or be affected by a proposed action, and sorting them according to their impact on the action and the impact the action will have on them. There is a mass of information on stakeholder analysis on the internet and in other sources so the topic will not be covered in detail here. However these are some key tips to apply when preparing for meetings with key senior executive stakeholders:
15
TRILLIUM SOFTWARE
Identify which senior executives are most important to your initiative and address
these as top priority.
Before meeting any of the senior stakeholders use your lower level contacts, the
corporate intranet and any other sources to try to build up a preparatory pen picture of the stakeholder. As a minimum you should know:
Their role within the overall organization. What areas do they lead? How
large? What current change programs do they sponsor? The last mentioned question is useful to see if your DQ work can support a current change program close to the executives heart.
Their organizations purpose, objectives and goals. How does your DQ

proposal relate to these objectives? How might poor DQ prevent the organization from achieving its goals? How will improving DQ make success more likely?
Their personal current priorities and hot issues what is currently keeping
them awake at night? What do they really care about? Try to put yourself in their shoes and think about the world as seen in their eyes.
Something about what type of people they are, what makes them tick and
what approaches are likely to find most favor in their eyes. A particularly useful technique for helping you to do this is attached at Appendix 2. Called Behavioral Styles Analysis the technique categorizes people into certain dominant behavioral types. Though inevitably a simplification it will help you to prepare your pitch at the meeting.
Once this analysis is done, decide who could / should act as the senior sponsor/
mentor for your DQ initiative. In some cases this sponsor will be self-selecting, for example the company CFO if the DQ project is focused on improving accounting or billing data. In cases where the answer is not obvious consider who is suffering the most pain as a result of the DQ issues you are proposing to address, and see them first.
16
TRILLIUM SOFTWARE
Meet with this senior executive and gain his / her support before moving on to other
senior stakeholders. Be very clear beforehand what you want this potential sponsor to do for you and what you will do for them. Examples of what they might do for you could include the provision of seed corn funding to help you kick start your DQ work, help you find business owners for data domains, identify key change agents in his or her organization whom you can work with, and to help you gain access to other senior executives. What you can do for them will of course include support to achieve their objectives, improvements to the bottom line of their organization, better data and hence more efficient processes, and so on.
Prepare for the meetings by sending the senior executive a statement of the purpose
of the meeting and an agenda. In this tell them what you think is in it for them. Use this to forward any supporting documents you want to refer to in the meeting.
In preparing material for the meeting you can choose a number of options. A good
way is to prepare a short briefing paper or presentation as this can help ensure you get your main points across in the (probably) short time you have been given. In Appendix 3 there is an example four slide presentation aimed at the CEO of a fictional hotel chain, though its format and style have roots in actual senior executive encounters. When preparing the briefing material remember to:
Express all potential benefits in bottom line terms. Ideally these should be
quantifiable and usually financial, though might also include brand reputation, regulatory compliance etc.
Highlight the risks to the senior executive of inaction. Showing him or her
that doing nothing is a high risk option will help to grab attention.
Think not only about the current business impact of DQ shortcomings

but also their potential threat to realising future goals and aspirations. Remember senior executives are paid to anticipate and plan for the future and not just manage the here and now.
17
TRILLIUM SOFTWARE
When answering questions use concepts and language that the senior executive
can relate to. Avoid DQ and technical jargon. Spell out the challenges and potential solutions in business terms. This will demonstrate that you know what his or her business is trying to achieve and how your DQ work relates to this. A great way of doing this is to develop DQ stories positive and negative which will bring what you are saying to life. A real life example of how poor DQ has impacted a senior executives area is worth a multitude of theoretical suppositions.
Be aware of DQ successes achieved in other organizations. This will enhance your credibility
by showing that DQ is an industry wide, global problem and that other organizations are tacking the same issues. Keep up to date with whats going on by keeping in touch with relevant websites, subscribing to DQ journals and magazines, and attending DQ conferences. If possible bring in an outside DQ specialist who has a track record of successful DQ delivery as its often easier to influence your senior executives with external corroboration. Having worked through the above, you should now be ready to meet your senior executives. Now we can move onto how to ensure success in the meetings themselves.
Step 5 Influence the Outcome

Wikipedia gives two complementary definitions of influencing as convincing others to take appropriate action and an action exerted by a person or thing with such power on another to cause change. In your meetings with senior executives this is key. The outcome you seek is to get them to support and become actively involved in your DQ initiative. Lots of information is available on the internet and elsewhere on improving your influencing skills so this will not be covered in detail in this white paper. Its well worth reviewing some of this in preparation for your meetings.
18
TRILLIUM SOFTWARE
There are a few key points about influencing that are worth highlighting here:
You will be more influential if you have carried out the preparation suggested in Steps 1 to
4 above. Sound preparation and research will make you more confident and self- assured. This will transmit itself to the senior executive and increase his or her confidence in you.
There is no single best way to influence others to achieve your desired outcomes. This
relates to the points made in Step 4 about customizing your preparation to best take account of the individual motivation, drivers and personality of the senior executive you are trying to influence.
Be flexible and prepared to compromise. You may not get everything you hope to in the
senior executive meetings. Some may support you strongly; others will be more guarded and sceptical. Demonstrating a willingness to make adjustments will show you are listening and open to advice. Though not what you may initially want this can actually help you to be more influential with that senior executive in the long term. Here are some wider tips from the authors own experience and from others who have trodden this path before:
Dont ask for too much, especially in initial meetings. Remember that the senior executive
will have lots of competing demands to juggle. This is best exemplified if asking for financial backing. You are much more likely to get a 50k investment to deliver a contained DQ project than 1 million for a major enterprise wide DQ initiative. Seek funding incrementally. Use early projects to make the case for further investment, and seek this once trust and a delivery record is established.
19
TRILLIUM SOFTWARE
Ensure you manage the senior executives expectations. The DQ problems you will
discuss have taken many years to develop and may take a long time to alleviate. Yet senior executives will expect quick results. Its therefore vital that your initial proposals contain initiatives that can deliver early. In general the golden rule is to under-promise and over-deliver. Success should then be seen by them as exceeding expectation, not a disappointment. Similarly be open about risks, and tell the senior executive what the risks are for him / her and you, and lay out your risk mitigation strategies. The risks of DQ work can at times be high, but so are the potential benefits.
If you are not experienced in dealing with senior managers initial encounters can be daunting.
It will be less intimidating if you are accompanied by a colleague, either from your own team or better still from a business unit within the senior executives own area. They can help you handle tricky questions and assist with note taking. Capturing a true record of what was agreed is important so making accurate notes is not a trivial matter.
There will be some encounters which do not go well. However sound your preparation and
assured your performance in the meeting you will not always get the desired outcome. If this happens analyze what went wrong and learn from it, but it may not always be your fault. Your executive might be having a bad day and its your ill fortune that you met them on it. If this does happen it can be difficult to repair. You can suggest another meeting where you can amend your pitch and try again. Alternatively use a friendly senior executive to try to reopen the door. As a worst case scenario think about damage limitation. Is this senior executive critical? If so, how can you counter / overcome any barriers he / she might put up in future? In many cases those reluctant to support you may come on board later once the early adopters have helped you to demonstrate its worth.
In most organizations attempts to improve DQ have been tried before. Some of these might
have failed, and some senior executives might remember these. This may make them cynical or resistant to your proposals. Ensure you are aware of these past attempts, have captured the lessons wherever possible, and show them why and how its different this time.
20
TRILLIUM SOFTWARE
At the end of the meeting always try to leave the door open for follow up meetings. Tell the
senior executive you will send them notes of what has been agreed. Do this within 24 hours. Also ask them how they want to be kept informed of progress going forward.
Last and not least, be confident. When it comes to DQ you are the expert, so act like a
leader. Who better than you to show leadership in DQ? because he wants to do it. As President Dwight D. Eisenhower observed, Leadership is at art of getting someone else to do something you want done
Step 6 Sustain Involvement

Once youve gained the senior executives initial commitment it is critical to sustain it as DQ improvement is a long term haul. Its no good planting the seed if you do not propagate the relationship and make it blossom. There are several ways you can do this:
Create organizational and data governance structures which give key senior executives a
continuing role. Advice on how to do this is explained in detail in a sister white paper entitled Organizing for Enterprise Data Quality Improvement which suggests how to involve senior executives in an organizational wide DQ improvement program. If your DQ project is less expansive than this the same rules apply. If the senior executive has a direct stake in the work he / she will remain more committed and improve your chances of success.
Whether you need senior executives to remain formally part of your work or not you have to
be able to demonstrate that you are delivering the promises you made to them. Producing regular, focused communications aimed at senior executives is vital. These could take many forms, including regular or ad hoc e-mail updates, webinars, further face to face briefings, or internal conferences / events. If organizing events, ask a senior executive to chair; this also helps to attract other executives in the organization and can encourage them to get involved. Remember that senior executives are busy people, so keep any communication short, business focused, in business language, and relevant to their direct interests and needs. Emphasise the benefits of what you have delivered, not how you did it.
21
TRILLIUM SOFTWARE
Ask for their advice and help when you need it, for instance to overcome resistance, to get
more cooperation from their people, and to help you understand and react to a change in business strategy or operation. Dont be a thorn in their side by disturbing them too often as this can be counterproductive but do remind them that the success of your initiative depends heavily on them.
Summary
Senior executives are not a breed apart. They can seem that way if we dont understand and share their dreams and visions. But in that respect they are no different from the rest of us. If others empathize with our dreams and visions and show us they can help us achieve them wed listen too. This paper has tried to help those who are trying to gain commitment via a simple six step set of actions. Following these will greatly increase your chances of opening the right doors and keeping them open. If DQ professionals and others intent on making improvement happen do their homework, demonstrate self-belief and act as leaders and not followers they can sell DQ to senior executives. If handled in the right way they can become your best allies. So reach for the top. It could be the springboard of your enhanced, high profile career as a DQ hero within your organization!
A new leader has to be able to change an organization that is dreamless, soulless and visionless someones got to make a wake up call.
Organizational Consultant
Warren Bennis,
22
Organizing for Enterprise Wide DQ Improvement

The experience of BTs Enterprise Information Quality Improvement Program
Harte-Hanks Trillium Software www.trilliumsoftware.com Corporate Headquarters +1 (978) 436-8900 trilliumsoftware@trilliumsoftware.com EMEA +44(0)118 940 7600 Central Europe +49(0)7031 714756
TRILLIUM SOFTWARE
Introduction
Data quality has matured. In its early days the focus of efforts to improve the quality of data in organizations was predominantly tactical. Usually a specific data quality (DQ) problem was identified and a project initiated and delivered to resolve or ameliorate it. Examples include improving a customer marketing list, clearing redundant records of former customers, matching logical and physical inventory and so on. This approach was characterised by a heavy emphasis on data cleanse, a one off process where shortcomings were recognized, quantified and improvements made. Although many organizations reaped rewards from this approach, often the underlying causes of DQ problems were at best partially addressed and sometimes not tackled at all. The end result was that data cleanse became a regular, reactive, routine activity, with some data sources cleansed again and again. All too often the DQ improvement achieved was not sustainable. Moreover these tactical approaches failed to recognize a critical truth about DQ, that the places in the organization where the problems were most acutely felt were often not the places where the problems originated. For example incomplete or inaccurate capture of new customer names and addresses by Sales may not have a major impact on Sales, but would undermine the efficiency of post sales activities including invoicing, delivery and product or service support. These shortcomings have led many organizations to move away from this tactical focus. Instead they have recognized the pervasive nature of DQ problems, and how an entire organization needs to be mobilized to fix them. To achieve this they adopt more strategic approaches where truly sustainable DQ improvement results from cross-organizational collaboration and interworking. To make this happen enterprise wide approaches are put in place, including organization wide improvement Programs, pan-organizational data governance and Master Data Management (MDM). This approach is not just quantitatively distinct from tactical initiatives but involves a radically different way of tacking DQ.
TRILLIUM SOFTWARE
This paper is aimed at all involved in DQ improvement, whether primarily in business or IT roles, who are considering, initiating or actively involved in an organization wide approach to DQ improvement. It draws on the authors experience in initiating and leading an enterprise wide DQ Program in a major global organization, British Telecommunications plc. (BT) and on consultancy engagements with global organizations and UK government departments. In 1998 BT started its Information Quality Improvement Program (IQIP). This organization wide Program ran for 10 years until 2007 when DQ improvement initiatives were absorbed into business as usual activities. When it ended IQIP had delivered over 75 DQ improvement projects, ranging widely in scope and purpose and affecting every BT line of business. IQIP had a significant impact on BTs bottom line. All projects were supported by business cases and these cases were evaluated at the end of each delivery. Overall more than 625 million of verified cost reductions and other benefits were realized. Today many of these initiatives are still in place and continue to help BT maintain its quality of data. The chief legacy of IQIP is that it played a fundamental role in changing the culture of the company to one where DQ is recognized as an essential prerequisite of an efficient business. The Program was praised by external market analysts. A 2006 Gartner report stated BT is one of the few companies that is actually meeting the challenge of managing DQ effectively. Further reviews conducted by Forrester, the Butler Group, and the University of St. Gallen Switzerland endorsed this assessment, This white paper highlights how BTs experience of IQIP can help to inform the challenge of organizing for enterprise wide DQ, lists the key organizational structures, outlines the roles of the potential main protagonists and summarizes the main lessons learned.
TRILLIUM SOFTWARE
Enterprise DQ Improvement
Improving DQ across a major business is a complex challenge. This inherent complexity is rooted in the fact that DQ is first and foremost a business change / transformation problem, not an IT challenge as it is sometimes incorrectly portrayed. DQ problems occur in organizations for a variety of reasons. These causes relate to people, process and technology. Examples include:
People
Human error Inadequate ownership of / accountability for data and its management Poor training Information as power internal politics Denial of DQ problems
Process
Poor process design Process failures Setting & enforcing inappropriate goals and objectives Absence of agreed data standards in business processes and IT systems
Technology
Data capture & update failures Multiple, usually inconsistent data sources Systems interface errors and omissions
E
PR
PE OP L
OC ES S
DATA QUALITY
TECHNOLOGY
TRILLIUM SOFTWARE
In most instances poor DQ arises from a combination of several factors occurring together, involving end to end shortcomings in people, process and technology. As such it is a systemic problem (meaning the whole is more complex than the sum of its parts) and so requires holistic approaches to improve it. This rule applies equally to small scale, tactical projects designed to tackle specific DQ issues all the way to major, strategic Programs of activity which extend right across an organization, for example Master Data Management (MDM) initiatives. A key goal of any enterprise DQ Program is to recognize this complexity and deal with it by breaking down the problem space into manageable and addressable chunks. It also implies that DQ problems can be improved, but not always resolved. Managing key stakeholder expectations is therefore also critical.
Implications for Enterprise Wide DQ Improvement Programs

Bearing these implications in mind any enterprise wide DQ improvement initiative should adhere to these ten maxims:
1. The overall initiative must be led by the business, and supported by IT. DQ improvement is
not an end in itself but an enabler to better business performance. Its potential benefits must therefore be recognized by the business, and owned by them.
2. Ensure that the Program has a senior business champion, ideally at CxO level within your
organization. He / she will promote and support the Program at the highest levels, help to publicize success and break down barriers that may (and probably will) be put in your way. Ideally this manager should sit within a central enterprise wide function to ensure neutrality and impartiality.
TRILLIUM SOFTWARE
3. The Program must gain the active support and involvement of all parts of your organization
(and potentially other third party organizations, e.g. IT providers, customer groups etc.) who are affected by DQ problems and who need to play a role in improving them. Early stakeholder analysis is essential, and ways of engaging with key stakeholders clearly mapped out. Given the complexity of larger organizations, and the potentially large number of affected stakeholders, this is a compelling reason for starting small, delivering early, and growing the Program incrementally, using early success as a foundation for more ambitious projects.
4. A common characteristic of DQ problems is that the place in the organization where they
are most acutely felt is often not where the problems arise, e.g. Sales create addresses when taking orders. If these are inaccurate or incomplete this may not be a major problem for Sales but will have a negative impact on Delivery. Fixing the problem therefore requires Sales and Delivery to work collaboratively. A key objective of any DQ Program at the enterprise level is to foster cross-organizational and cross-functional collaboration so the Program must give this a high priority.
5. Specific DQ projects generated by the initiative must in every case have a nominated
business owner / champion. This business owner must be responsible for identifying the business benefits of successful delivery and ensuring they are realized.
6. No DQ project should be started without a signed off business case. Business cases are
best generated on a per project basis, rather than trying to create an overriding business case for the Program as a whole. This is because every project will address a unique combination of people, process and technology issues and so its benefits and costs will be specific to that project, e.g. the benefits of cleansing a client address file are very different from those of an initiative to ensure regulatory compliance in a reference data file.
7. Walk before you can run. Early projects must succeed otherwise the entire Program will be
jeopardized. Prioritization and delivery of these initial projects are critical to prove the value of the Program as a whole and to demonstrate early success to senior sponsors. They also instill a sense of urgency.
TRILLIUM SOFTWARE
8. Recognize the DQ maturity of your organization. If your organization is one where DQ is

not recognized as a significant problem by most people, it is highly unlikely that proposals to deliver expansive, expensive, all embracing improvements (e.g. enterprise wide MDM) will win favor. It is vital to evolve your Program, starting with acknowledged specific DQ problems, delivering improvements, proving your approaches, and using these to make the case for more strategic support & investment.
9. Given the scale, scope & complexity of data challenges at the enterprise level it is essential
to develop common approaches, methodologies and DQ toolsets. These will ensure that solutions are reusable, expandable and sustainable. DQ tools are essential to support data analysis and data re-engineering activities.
10. Be contingent and flexible. The Program will need to change and evolve to meet changing
business drivers, business reorganizations, replacement of key stakeholders and so on. Be prepared to adapt your Program to cater for these changes, and to adjust the Program to meet changing demands.
Getting Organized the BT Experience

The fact that an enterprise wide DQ improvement Program has been initiated is a clear indication that current ownership of and accountability for DQ problems within an organization is at best less than adequate or even non-existent. Creating a new organization to tackle DQ problems is therefore a key challenge for the Program. Organizing for enterprise wide DQ improvement should take account of the ten maxims above. The exact organizational shape and constitution will vary from organization to organization depending on the local characteristics of the enterprise. Nevertheless, based on BTs experience with IQIP, a generic proposed structure follows:
TRILLIUM SOFTWARE
PROPOSED ORGANIZATION STRUCTURE FOR ENTERPRISE WIDE DQ IMPROVEMENT PROGRAM
EDQ Steering Group
EDQ Program Board
Business Area A Board
Business Area B Board
Business Area C Board
Project Board 1
Project Board 2
Project Board 1
Project Board 2
The composition, roles and functions of each of the main structures (Steering Group, Program Board, Business Area Board and Project Board) is summarized below:
1. EDQ STEERING GROUP

Chaired by the senior business manager who is the designated champion of the initiative. Other members should be senior business representatives from all main stakeholder areas identified in prior stakeholder analysis. These would normally be representatives of all main business areas, plus a senior IT manager (ideally a CIO) and senior representatives of other important participants, for example third party suppliers. The EDQ Program Lead (who chairs the EDQ Program Board) should also be a member and acts as the primary link between the Steering Group and Program Board.
TRILLIUM SOFTWARE
The main roles of the EDQ Steering Group are to:
Provide overall governance of the Program Ensure the Program aligns with, and supports, the strategic priorities of the organization so
that future as well as current issues are addressed
Help the Program identify business hot spots where focus on DQ could make the biggest impact Secure commitment and investment from across the enterprise Resolve cross-organizational issues that are flagged up from the Program Board Endorse the prioritization of specific DQ projects Sign off funding of business cases & individual projects Monitor overall EDQ Program progress Communicate progress and benefits to the main Board of Directors
From BTs experience members of this board must be empowered to take decisions on behalf of the area they represent. Personal commitment to the EDQ initiative is more important than their formal position within the overall organizational hierarchy though they must be senior enough to secure the necessary commitment. Generally this Board needs to meet at least quarterly, possibly monthly in the early stages of the EDQ initiative.
2. EDQ PROGRAM BOARD

Chaired by a full time Program Manager who also sits on the EDQ Steering Board, this Board manages the day to day running of the EDQ Program. Like the Steering Group chairperson, ideally the incumbent should be part of a business unit at group level. Other members should consist of senior or middle managers from the various business areas and other stakeholder groups involved in the Program. It should also contain the primary IT Program Manager responsible for overall delivery of all technical components of the Program. This board would normally meet monthly. In BTs Program Board (called the Information Management Forum) two additional roles which proved of great value were:
A finance specialist, whose primary job was to work with others to build, write and endorse
business cases for specific DQ projects.
TRILLIUM SOFTWARE
A communications manager, who ensured successes were publicized up and across the
organization, produced case studies, and generated marketing material to promote the merits of the EDQ initiative. The main roles of the Board are those that could be expected of any Business Program Board and include:
Set and communicate a clear vision for the Program Program planning and control Definition, implementation & monitoring of KPIs / CSFs Program resource management & skills development Program risk and issue management & mitigation Benefits management endorsing business cases and ensuring deliveries realize the projected benefits Stakeholder management Prioritization of current and proposed DQ projects Financial control of the Program Managing project interdependencies where projects embrace more than one business area Selection, implementation and enforcement of Program wide processes, methodologies & IT DQ tool sets
Regular and exception reporting to the EDQ Steering Group
3. BUSINESS AREA BOARDS

BT is a large, complex global organization with a turnover of 20 billion+ per annum and 90,000+ employees worldwide. Once its IQIP Program had expanded and accelerated it was decided to create sub-Program boards as the pace and volume of DQ projects were putting undue pressures on the original Program Board. In BT these were based on its three primary customer facing organizational units (termed lines of business), which at the time of IQIP were BT Retail, BT Wholesale and BT Global Services.
10
TRILLIUM SOFTWARE
Each Business Area board was chaired by the business lead who represented his / her line of business on the overall EDQ Program Board. Other members consisted of business and IT managers who were responsible for the delivery of business DQ improvement projects led by that line of business (LoB). Note in BTs case all DQ projects were eventually designated a primary line of business owner as it was deemed that a single LoB should own the business case, even if other LoBs were actively involved and would benefit. Consequently all individual DQ projects were managed by the appropriate Business Area boards, allowing the EDQ Program board to focus on more strategic matters. Business Area boards normally met monthly. The roles of the Business Area boards were primarily as those of the Program Board for the elements of the sub-Program they led. One key role in the BT structure is worth highlighting as it proved to be pivotal to the success of IQIP. The IT department created three DQ Consultancy roles. Each role faced a specific line of business. These consultants were DQ experts, acquiring hybrid business and IT skills. They developed a close understanding of the LoB business drivers and the impact on these of DQ. As a result they were able to help their LoB identify and evaluate potential DQ improvement projects, develop business cases, define and design improvement project proposals and work with IT to design and deliver the projects.
4. PROJECT BOARDS
Normally chaired by a member of the Business Area board, they conducted all project management of specific DQ projects. In most, though not all cases see below - they also contained the lead IT project manager. They would normally meet at least weekly. All DQ projects need to be subject to the usual disciplines and practices of project management and so roles were as would be expected of any project management board. These are generic and so not further described here. Overall these organizational structures are advisory only. The optimum structure for any organization is contingent upon the specific culture, drivers and structures of that organization.
11
TRILLIUM SOFTWARE
Other considerations & learning points

1. ORGANIzING IT DELIVERY
In its IQIP BT used a combination of in house and third party IT suppliers. In the initial stages of IQIP IT resources were distributed across several organizational units. It soon became evident that this was a sub-optimal model as lines of communication were complex and it was difficult to reassign resources to meet changing priorities. As a result a case was made successfully to re-organize IT delivery. All IT resources working on the IQIP Program were reassigned to a single IT unit. This became the Information & Knowledge Management Centre of Excellence (IKM CoE). The head of IKM then became the obvious senior IT manager to sit on the DQE Program board. This centralized IT organizational model had several significant advantages:
Dedicated focus on IQIP and related DQ projects meant IQIP did not have to contend for
scarce resources
Single IT point of contact established More flexible and simplified IT resource planning & allocation DQ expertise and skills development was centralized and so easier to manage and develop;
mentorship and apprenticeship was encouraged and expected
Enhanced ability to reuse approaches & solutions Third party resource management was simplified Easier to mandate common tool sets for DQ improvement Software licensing was rationalized and made more cost effective
12
TRILLIUM SOFTWARE
For delivery of the IT components of business DQ projects the following roles / skills were required:
DQ consultants (see earlier) Business / data analysts. These were the key users of data profiling & data modelling tools Technical architects / solutions designers. Their main role was to design the IT components
of DQ projects
Project managers with an understanding of DQ and its specific requirements. Software designers and developers with particular expertise in one or more of data profiling,
ETL, data migration, metadata management, BI, DBMS or DQ re-engineering (e.g. expertise in Trillium Software Quality) In BTs experience around 95% of DQ improvement projects required IT changes. In the other 5% of cases projects required no IT involvement but were exclusively focused on process changes and improvements, better training, improved staff controls & incentives to improve DQ as part of their everyday jobs etc.
2. DATA GOVERNANCE & DQ IMPROVEMENT

Data governance is currently a hot issue in DQ. It centres on the need to ensure all key data within an organization is actively managed and improved by making individuals within the organization accountable for subsets of that data. There are three basic alternative models of data governance:
Data centric: often termed data stewardship. Here business appointed people are tasked
with tending and enhancing key data domains wherever those domains are created or used across the organization. For example a named individual is responsible for all customer contact data across an organization.
Process centric: business process owners become the data owners for all data created,
amended or deleted by the business process for which they are responsible.
Systems centric: business IT system owners become the data owners for all data created,
amended or deleted by the system(s) they own.
13
TRILLIUM SOFTWARE
In BT all these models were employed at varying times and in different lines of business during IQIP as a way of ensuring the DQ gains made were maintained. All were found to have advantages and disadvantages. Which is the best for any organization depends on the particular needs of specific business areas. It is also worth emphasising that different governance models can be applied at the DQ project improvement and business as usual phases as DQ becomes an integral part of business as usual activities. Key is that accountability is placed on individuals who are committed to DQ improvement and have the time and wherewithal to make it happen.
Conclusion
Persistent problems require persistent solutions. To tackle persistent and pervasive data quality problems organizations need to embrace and implement approaches which are permanent, transformational and radical. The days of point solutions have had their day and need to be superseded by all-encompassing new approaches where data quality improvement is part of the lifeblood of organizations. This paper has tried to help all those who wish to embark on this path or are already on the journey. In the words of the philosopher Alan Watts The only way to make sense out of change is to plunge into it, move with it, and join the dance.
About the author:

Nigel Turner is an independent consultant and writer on data management. Prior to this he spent much of his career in British Telecommunications plc. (BT) where he led an enterprise wide data quality improvement Program, provided data management consultancy to several of BTs major customers, and ran a successful Information Management and CRM practice. He has published several papers on data management and is a regular invited speaker at CRM and Information Management events. Recently he was a co-author of the Institute of Direct Marketings online Essentials of Data Management training course. In 2007 he was voted by fellow data management professionals as runner up in Data Strategy magazines UK Data Professional of the Year.
14

Data Governance Ebook Final

Загружено:

Сведения о документе

Исходное описание:

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Data Governance Ebook Final

Загружено:

Авторское право:

Доступные форматы

A De nitive Guide

What Is Data Governance?

Understanding Data Governance

Scoping the Data Governance Initiative

Selling Data Quality to Senior Executives

Data Quality versus Data Governance

Organizing for Enterprise-wide Data Quality

Leveraging Technology for Data Governance

For more information

Understanding Data Governance

The Reason for Data Governance

Fitting Data Governance into an Organization

Trillium Software ______________________________________________________________________________

Common Oversight Body/Organizations for Company Assets

Order to cash processes

Human Resources Data Governance

Leading Data Governance within an Organization

Scope of a Data Governance Program

Trillium Software ______________________________________________________________________________

Best practices and policy around data entry, validation, access

Business Rules Data Architecture Technology

Data processing, transformation, formatting, cleansing, matching

Data modeling and integration

Data quality tools, MDM, CRM, ERP, etc.

Table 2. Data Governance Domains/Disciplines

Trillium Software ______________________________________________________________________________ Data Governance Drivers

Benefits of Enterprise Data Governance

Trillium Software ______________________________________________________________________________ Who Should Consider Data Governance?

Mission-Critical Best Practices for Data Governance

Trillium Software ______________________________________________________________________________

(978) 436-8900r (719) 481-0565

Reach for the Top

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

Why should senior executives care about data quality?

In a 2009 Information Difference survey of global organizations with turnovers of $1 billion or

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

Why are senior executives important to data quality?

As a business transformation challenge DQ improvement requires champions and mentors

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

So why is it so hard to engage senior executives?

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

Finally DQ professionals must shoulder much of the blame. Many DQ specialists in

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

The Solution How to Sell Data Quality to Senior Executives

STEP 1 - ARTICULATE THE PROBLEM

STEP 2 CREATE THE VISION

STEP 3 DEMONSTRATE DELIVERY

STEP 4 - PREPARE THE GROUND

STEP 5 - INFLUENCE THE OUTCOME

STEP 6 - SUSTAIN INVOLVEMENT

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

SIX STEPS TO THE TOP: Selling Data Quality to Senior Executvies

Influence the Outcome

Prepare the Ground

Create the Vision

Articulate the Problem

Step 1 Articulate the Problem

Copyright 2011 Harte-Hanks Trillium Software Alll rights reserved www.trilliumsoftware.com

DATA QUALITY SYSTEMIC FAILURE

Human error No data accountability Poor training Internal politics Denial

This diagram demonstrates that: