Executive Summary
Data Management is a critical aspect of organisational operations, involving various stakeholders such as data citizens, professionals, and executives. ‘Reference & Master Data Management for Data Managers for Data Managers’ encompasses understanding and managing Master Data, data quality, governance, and challenges in data integration and sharing, as well as the role of Master Data in data efficiency. Howard Diesel covers the revised version of the DMBoK version 2, the implementation of Reference and Master Data and the impact and management of referential data errors. The webinar highlights the importance and process of implementing Reference and Master Data in an organisation, the role of data in business transactions, and the significance of Data Governance and Master Data Management. Additionally, Howard covers the key stakeholders and RACI models in Data Governance and the concept of golden values, records, and profiles in Master Data Management.
Webinar Details
Title: Reference & Master Data for Data Managers
Date: 12 September 2024
Presenter: Howard Diesel
Meetup Group: African Data Management Community
Write-up Author: Howard Diesel
Contents
Second Version of the Master Data Management Course
Understanding the Role and Responsibilities of Data Personas in the Organisation
Sample Questions in Data Management for the CDMP Specialist Exam
Understanding and Managing Master Data
Challenges of Data Management for Master and Reference Data Initiatives
Data Management and Analysis Techniques
Understanding What has been Revised in DMBoK Chapter: Core Elements, Definitions, and Implementation
Data Quality, Governance, and Challenges in Data Management
The Process of Implementing Reference and Master Data in an Organisation
Data Quality and Governance in Data Integration and Sharing
Data Quality and the Role of Master Data
Golden Records in Data Warehousing
Data Management in Data Vault Warehouses
Key Stakeholders in Data Governance
Understanding Golden Values, Records, and Profile in Master Data Management
The Terminology and Integration of Systems of Record in Master Data Management
Importance and Impact of Different Data Types in Business Operations
Understanding the Role of Data in Business Transactions
Transaction Data and the Role of Master Data in Data Management
Impact and Management of Reference Data Errors
Data Governance and Readiness Assessment in Organisations
Understanding Decision Mandates and Data Architecture in MDM
The Importance of Data Governance in Master Data Management
Master Data Management and Data Governance
Second Version of the Master Data Management Course
The webinar opens with Howard Diesel explaining that this will be the second version of the Reference and Master Data Management course and that Modelware Systems will be refreshing all courses to align with the revised edition of the DMBoK. Although there aren’t many changes, it is important to ensure that the wording and terminology are used correctly and align with the revised material.
Understanding the Role and Responsibilities of Data Personas in the Organization
The webinar series will cover four main roles related to Data Management: data citizen, data professional, data manager and data executive. Data citizens are responsible for using and managing data within the organisation while adhering to policies and procedures. They create standard operating procedures for Master Data and Reference Data. Data professionals are in charge of data models and work with various knowledge areas such as data architecture, data integration, and data quality. The series will also address the challenges of building a business case for Master Data, choosing between operational and analytical approaches, and aligning Data Management with the overall business strategy, including subject area ownership and governance framework allocation.
Figure 1 Master & Reference Data Management Webinar Series
Sample Questions in Data Management for the CDMP Specialist Exam
Howard touches on preparing for the CDMP Specialist Exam by delving into detailed areas and understanding the type of questions to be expected. Sample questions are presented, prompting participants to share their thoughts on the common motivation for Reference and Master Data Management. Howard engages with the participants’ answers, acknowledging their responses and emphasising the importance of identifying the best answer rather than a correct one as this relates to the Specialist Exam.
Figure 2 Sample of Questions and Answers
Understanding and Managing Master Data
The definition of Master Data and Reference Data are brought to the attendees’ attention. Master Data refers to data about business entities that provide context for transactions. In contrast, Reference Data is data used to classify or categorise other data and is widely accessed and Referenced across an organisation. The primary areas of focus in Master Data Management (MDM) include generating “a golden record,” identifying duplicate records, producing read-only versions of key data items, providing access to golden data records, and producing clear data definitions for Master Data.
Challenges of Data Management for Master and Reference Data Initiatives
Howard emphasises the importance of pursuing Reference and Master Data initiatives, highlighting the role of Master Data initiatives as essential functions in the Data Management framework. Centralising the management of Reference and Master Data creates job security for data professionals and allows them to conform critical data needed for analysis. Howard raises questions about the similarities and differences between managing Reference and Master Data and the techniques for splitting or merging instances of a business entity.
Data Management and Analysis Techniques
The webinar explores the classification of different types of data, including Master Data, Reference Data, and read-only versions. Centralised management is imperative for Reference and Master Data as it ensures critical data availability for analysis. The concept of “survivorship” is explained in the context of merging records and building the golden record.
Understanding What has been Revised in DMBoK Chapter: Core Elements, Definitions, and Implementation
The DMBoK chapter covers core elements such as managing reconciled and integrated data through stewardship and semantic consistency in support of enterprise-wide data-sharing needs. The definition and deliverables have been revised in the latest edition. Different architectural approaches, such as virtualised, registry, centralised hub, and transaction hub, can be chosen based on the specific domain needs. Recognising that not all domains require the same approach and technique is important. Virtualised approaches are becoming more common, allowing data transformation without data duplication. Modelling the data is necessary when using a product without existing models, enabling the identification of expected elements and differences. Establishing policies and implementing data sharing and integration services are also essential components.
Figure 3 Chapter 10 – RMD DMBOK Overview
Figure 4 What is Reference & Master Data?
Figure 5 Reference & Master Data Activities
Data Quality, Governance, and Challenges in Data Management
The importance of data quality in the context of integrating various records cannot be overstated. It involves identifying duplicates, cleaning, and standardising the data, which are essential deliverables within the environment. Governance plays a significant role in recognising and integrating data sources, requiring a data catalogue to locate the relevant information, particularly in dealing with customer data across different sources. The challenge of dealing with duplicated data due to silos, and the need to identify the system of record or master application, adds another layer of complexity. Data quality rules, conditions of use, and monitored activities are crucial, as allocating business procedures and resolving matching processes may require time and additional communication with customers. Additionally, meeting stakeholder needs involves presenting comprehensive information and ensuring standard approval gates, expectations, and deployment of Reference and Master Data are in sync, which forms a critical part of the business case and addresses current challenges. The overview provided aligns with the discussions related to the Data Management box.
Figure 6 Reference and Master Data Governance
The Process of Implementing Reference and Master Data in an Organization
Key areas to focus on when managing Reference and Master Data include consolidating data from various departments such as sales, marketing, HR, finance, and IT and reducing duplicate copies of Reference Data. An attendee recommends centralising ownership and creating a demilitarised zone (DMZ) for the shared Reference and Master Data , removing politics and establishing a centralised authoritative source for the organisation. This approach enhances data security and streamlines data access for internal and external users, with the exact location of the centralised data dependent on the specific implementation style.
Figure 7 Reference & Master Data: Siloed
Figure 8 Shared Reference & Master Data BELONGS to the Organisation
Data Quality and Governance in Data Integration and Sharing
The importance of Reference Data in an ongoing data quality program and its role in data integration and achieving quality is explored. Challenges of implementing such projects, particularly within government entities, are highlighted due to the need to establish Data Management committees and models before moving on to MDM projects. Readiness checks, including the establishment of Data Governance and a culture that understands Data Governance, are crucial before implementing Reference and Master Data.
Figure 9 RMD is an Ongoing Data Quality Program
Data Quality and the Role of Master Data
Ongoing focus on data quality is crucial, especially in relation to Master Data. Ensuring consistency and accuracy across all organisational data is a top priority, with particular attention given to the central repository of critical data. The challenges of siloed data and the need for a consistent view across different applications are evident, highlighting the importance of addressing data quality issues. The role of Master Data or Reference Data systems in resolving these issues is imperative as they go hand in hand with data quality problems. Howard underscores the significance of data privacy by mentioning Informatica’s data subject master domain as a noteworthy approach.
Golden Records in Data Warehousing
Creating “a golden record” requires the elimination of redundancy from multiple data sources. An attendee inquired about the possibility of maintaining historical records while creating a single golden record. The aim is to ensure that all applications present the golden record while retaining the ability to access the original records if needed and whether it is feasible to access the original records after the match-link process without being aware of the merge.
Data Management in Data Vault Warehouses
The establishment of a “golden record” is for the purpose of ensuring consistency. Various techniques for establishing and maintaining the golden record were explored, including merging, linking, and deduplication, with an emphasis on the importance of a good MDM system for automating these processes. The impact of good MDM on downstream systems leads to more benefits of a centralised authoritative source of data for reporting and risk management. Additionally, the use of Data Vault as a collection area for raw data from various systems was mentioned, highlighting its role as a stepping stone for building a Master Data system. Additionally, the Data Vault’s ability to keep a satellite by system allows visibility into all system data ingested from any point in time and keeping it in a common area.
Key Stakeholders in Data Governance
The key stakeholders for our data strategy include the Chief Data Officer, responsible for the overall data strategy and governance; the Data Governance steering committee, accountable for funding initiatives; and the Data Governance Council, ensuring successful delivery of Reference and Master Data initiatives. Data owners are responsible for specific data domains, data stewards handle the day-to-day management of domain values, and the IT department is responsible for technical infrastructure and managing tools like Informatica and storage. These stakeholders play crucial roles in prioritising and implementing Reference and Master Data programs within the organisation.
Figure 10 Golden Values, Record, & Profile
Understanding Golden Values, Records, and Profile in Master Data Management
Howard mentions the concept of “golden values,” “golden records,” and “golden profiles” in the context of Master Data Management. “Golden values” refer to the most accurate and reliable values for specific attribute columns or sets before data integration. “Golden records” are comprehensive consolidated views of entities, while “golden profiles” provide a holistic view of an entity across various dimensions and attributes. The process of creating these involves data collection, cleaning, and integration to form an authoritative source for analytical purposes. Howard also suggests using analytical MDM initially and considering real-time updates or operational MDM based on specific challenges and requirements.
Figure 11 How do Golden(s) work together
The Terminology and Integration of Systems of Record in Master Data Management
The different terminologies related to Data Management, including system of record, system of Reference, and analytical system of record, are mentioned. To facilitate efficient Data Management, it is important to integrate and harmonise data from various operational systems into an analytical system of record. The process involves ETL (extract, transform, load) from applications to the Master Data and synchronisation to keep different systems up to date. Howard explores the concept of reverse ETL for pushing data back to operational systems and publishing to the enterprise data warehouse. Additionally, he highlights the significance of achieving analytical MDM before operational MDM due to the challenges associated with standardising diverse applications.
Figure 12 Systems Of … Landscape
Importance and Impact of Different Data Types in Business Operations
In Data Management, it’s crucial to recognise different types of data, such as Metadata, Reference Data, Master Data, transaction data, local data, and derived data. Understanding the impact of data errors at various stages is vital. Errors in Master or Reference Data can have far-reaching consequences, affecting downstream processes, while issues in transaction data may not have the same impact. This highlights the importance of maintaining accuracy and integrity at all data levels within an enterprise.
Figure 13 Types of Data
Figure 14 Impact of the different types of data
Understanding the Role of Data in Business Transactions
The data in the system can be categorized into Reference Data, category information, and Master Data, which provide high levels of abstraction for data modelling. This, context data, is used for business transactions and additional information. We also deal with transactional data, which includes business events, facts, dimensions, and data generated by business transactions. Furthermore, detailed transaction information helps create wide data sets and gather Internet of Things or sensor data that can be linked to transactions. For instance, when a customer purchases a double espresso at Starbucks, the currency used is Reference Data, the payment is a business event, and the details of the espresso are stored in the product Master Data. Additionally, detailed data or sensor data from the coffee machine can be analysed to address any complaints about the coffee quality. Various models and methods are used to collect and manage this data effectively.
Transaction Data and the Role of Master Data in Data Management
The transaction data involves breaking down event descriptions into who, what, where, why, when, how, and how many and further analysing the elements using customer Master Data. This process allows for a complete transaction view, with details such as product code, location, date, payment method, quantity, currency, and amount stored in the fact table using surrogate keys. Master Data and Reference Data complement transaction data by providing additional valuable information for advanced analytics, such as correcting assumptions about individuals and their associations. MDM systems can also handle slowly changing dimensions, reducing analytical processes and data storage complexity. Overall, leveraging Master Data systems enables the incorporation of detailed information, enhancing understanding and decision-making capabilities.
Figure 15 Putting it all together
Figure 16 What is Transaction/Event Data
Figure 17 Transaction/Event Data Example
Figure 18 About Transaction/Event Data
Impact and Management of Reference Data Errors
The next step involves addressing a Reference Data error and its potential impact, which can affect all work orders created against the assets. Unlike Master Data errors, which may only impact specific resources or transactions, Reference Data errors can have a more widespread effect, especially when there are incorrect definitions or data models. Managing Reference Data changes is complex and requires careful consideration to understand and communicate the potential impact and to execute updates effectively. A notable example is the currency change in Zambia, which led to significant disruptions due to a change in currency code and the reduction of decimal points, requiring extensive accounting adjustments.
Figure 19 Importance of Reference Data
Figure 20 Reference Data Change Management
Data Governance and Readiness Assessment in Organisations
When assessing readiness for Data Governance, consider factors such as maturity, culture, and the level of change within the organisation. Conduct a DMMA (Data Management Maturity Assessment) to evaluate the state of governance, quality, data integration, data modelling, and Metadata. Prioritise governance, risk understanding, employee engagement, collaboration, and accountability to drive data ownership and Key Performance Indicators (KPIs)-driven changes. Change management is crucial in facilitating the transition and addressing organisational concerns.
Figure 21 Readiness Assessment
Figure 22 Data Management Maturity Assessment Objectives
Figure 23 Corporate Culture Requirements
Understanding Decision Mandates and Data Architecture in MDM
During a discussion, the concept of decision mandates is emphasised, dividing them into big bets, cross-cutting decisions, delegated decisions, and ad hoc decisions. Howard outlines the involvement of different personas, such as the CDO steering committee, data managers, and professionals, in making, monitoring, and being responsible for various decisions. The importance of understanding dependencies, particularly in relation to data architecture, is highlighted to avoid implementation issues. Additionally, Howard encourages collaboration with the data architect to navigate these areas effectively.
The Importance of Data Governance in Master Data Management
Organisations face challenges when implementing MDM without proper governance structures, roles, and responsibilities in place. Pursuing an MDM program can serve as a catalyst for establishing these essential elements and building momentum for them. However, Howard notes that successful MDM implementation requires a strong focus on Data Governance and data quality and that leading with MDM without considering Data Governance can lead to failures. The importance of having a comprehensive data strategy, prioritising issues, and understanding technical feasibility before implementation was highlighted.
Figure 24 Organisational Change Management
Figure 25 RMD Decisions
Figure 26 RMD Manager Decisions
Figure 27 Integrated Business Capabilities & Data
Master Data Management and Data Governance
The importance of MDM as a driver for Data Governance is discussed. Howard mentions the need to have Data Governance in place to ensure the success of MDM and the idea that MDM can organically lead to the formation of Data Governance structures. Start small and prove the value of MDM before fully implementing it. Consider the organisation’s size and readiness. Additionally, having a data catalogue in place as a starting point for MDM practice and the potential of open platforms like Data Hub for building a data dictionary of different data assets are recommended. Howard closes the webinar expressing gratitude to the attendees.
If you would like to join the discussion, please visit our community platform, the Data Professional Expedition.
Additionally, if you would like to watch the edited video on our YouTube please click here.
If you would like to be a guest speaker on a future webinar, kindly contact Debbie (social@modelwaresystems.com)
Don’t forget to join our exciting LinkedIn and Meetup data communities not to miss out!