AVOIDING SILOED DATA AND SILOED DATA MANAGEMENT Dalton Cervo Author, Consultant, Data Management Expert March 2016 This presentation contains extracts from books that are: Copyright 2011 John Wiley & Sons, Inc. Copyright 2015 Elsevier Inc.
Synopsis Siloed, disparate, fragmented, and conflicting data are a fairly known issue faced by many companies today Companies have come to the realization Data Management disciplines are a must to address their data problems However, Data Management disciplines themselves cannot remain siloed Highest potential can be achieved by properly blending data management disciplines
About Dalton Cervo President and founder of Data Gap Consulting, providing data management consulting services in Master Data Management (MDM), Data Architecture, Data Integration, Data Quality, Data Governance, Data Stewardship, Reference Data Management, Metadata Management, Data Lifecycle Management, Data Warehouse, and Analytics & Business Intelligence. Over 24 years of experience in data management, project management, and software development, including architecture design and implementation of multiple MDM solutions, and management of data quality, data integration, metadata, data governance, and data stewardship programs. Experience in a wide variety of industries, such as automotive, telecom, energy, retail, and financial services.
About Dalton Cervo (cont.) Prior to Data Gap Consulting, served as a consultant for SAS/DataFlux, providing expert knowledge in MDM, data governance, data quality, data integration, and data stewardship. Prior to that, held the position of senior program manager at Sun Microsystems and Oracle Corporation, serving as the data-quality lead throughout the planning and implementation of Sun s enterprise customer data hub. Coauthor of the following two books: Multi-Domain Master Data Management Advanced MDM and Data Governance in Practice (Morgan Kaufmann, Elsevier, April 2015). Master Data Management in Practice: Achieving True Customer MDM (John Wiley & Sons, June 2011).
Publisher: Morgan Kaufmann Publication date: April 8, 2015 Multi-Domain Master Data Management delivers practical guidance and specific instruction to help guide planners and practitioners through the challenges of a multi-domain master data management (MDM) implementation. Authors Mark Allen and Dalton Cervo bring their expertise to you in the only reference you need to help your organization take master data management to the next level by incorporating it across multiple domains. Written in a business friendly style with sufficient program planning guidance, this book covers a comprehensive set of topics and advanced strategies centered on the key MDM disciplines of Data Governance, Data Stewardship, Data Quality Management, Metadata Management, and Data Integration.
In this book, authors Dalton Cervo and Mark Allen show you how to implement Master Data Management (MDM) within your business model to create a more quality controlled approach. Focusing on techniques that can improve data quality management, lower data maintenance costs, reduce corporate and compliance risks, and drive increased efficiency in customer data management practices, the book will guide you in successfully managing and maintaining your customer master data. You'll find the expert guidance you need, complete with tables, graphs, and charts, in planning, implementing, and managing MDM. Publisher: John Wiley & Sons, Inc. Publication date: May 25, 2011
Avoiding Siloed Data and Siloed Data Management
Agenda Data, Information, and Knowledge State of Affairs Avoiding Siloed Data with MDM Avoiding Siloed Data Management Final Thoughts
DATA, INFORMATION, AND KNOWLEDGE
The Basics Data 4, 2 (without context, these values are meaningless) Information Temperature 4 C, Dew Point 2 C (context adds meaning) Knowledge A temperature of 4 C and a dew point of 2 C, together with a rain, means that there is a chance of icing. This icing can adversely affect the performance of an aircraft. This is the same conditions that led to an accident last year. Deicing is recommended.
Structuring of Data Structured Data Semi-structured Data Unstructured Data
Categories of Data (main ones) Master Data Data representing key data entities critical to a company operations and analytics because of how it interacts and provides context to transactional data Transactional Data Data associated to or resulting from specific business transactions Reference Data Data typically represented by code set values used to classify or categorize other types of data such as master data and transactional data Metadata Descriptive information about data entities and elements such as regarding the definition, type, structure, lineage, usage, changes, and so on Others: Historical data, temporary data, etc.
Big Data (3 V s) 5 V s: + Veracity, Value
STATE OF AFFAIRS: TYPICAL COMPANY
State of Affairs 57% of all companies need more than two days to generate a complete list of customers. 75% of information workers have made business decisions that later turned out to be wrong due to flawed data. Up to 70% of IT resources are spent on building and maintaining connections between systems. A total of 56% of CIOs and IT managers could integrate less than 40 percent of their IT applications with other applications in their organization.
State of Affairs (cont.) Over the next two years, more than 25 percent of critical data in Fortune 1000 companies will continue to be flawed, that is, the information will be inaccurate, incomplete or [unnecessarily] duplicated The size of the digital record will grow by a compound annual growth rate of 60%.
Typical Company Internal Systems Vendors Supply-Chain Management Reference Data Management Operational Systems ODS Business Reports Other 3 V s Order Mgmt MDM CRM ERP EDW and Data Marts Analytics, Business Intelligence Big Data Management Social Media
How to solve the problem? Traditional, application-driven organizations Transform Data-driven organizations
Data Management Data Quality Management Data Governance Data Stewardship Metadata Management Data Integration and Synchronization MDM Data Management Reference Data Management Data Security Data Architecture Big Data Predictive Analytics
AVOIDING SILOED DATA WITH MASTER DATA MANAGEMENT (MDM)
Why MDM? Vendor Customer Product Employee Master Data Patients Sites Service Providers
The Narrow and Shallow View of Domain Data Col 1 Col 2 Col 3 Col 4 Col 5 Col 6 Col 7 Col m Row 1 Row 2 X X X X Row 3 Row 4 X X X Row 5 X X X Row n *Table represents the full set of data for a particular domain at a given source Business Process 1** Business Process 2** **Business processes use a small set of data
Business Case MDM gives companies the opportunity to better manage its key data assets and thereby improve the overall value and utility the data provides within the company It exposes internal process issues and business practices (or lack thereof) that are the underlying constraints to having and maintaining good data
Business Case (cont.) Lack of data management practices leads to: Increased costs due to operational and data redundancies or differences across lines of business. Higher risk of audits and regulatory violations. Poorer BI and analytics, adding to customer frustration and missed opportunities. Customer/partner/vendor/employee dissatisfaction and consequently un-realized revenues. Possible overpayment of vendors and customers stemming from duplicate records. Over or under delivery of customer services due to inconsistent customer identity and tracking.
MDM Business Case
Prioritizing Domains
MDM Styles Analytical Registry Style Transaction or Persistent Style Hybrid Style
Analytical MDM
Registry Style
Transaction or Persistent Style
Hybrid Style
Why is MDM Complex?
AVOIDING SILOED DATA MANAGEMENT
Data Management
Data Quality Management (DQM) DQM is about employing processes, methods, and technologies to ensure the quality of the data meets specific business requirements Trusted data delivered in a timely manner is the ultimate goal DQM can be reactive or preventive. More mature companies are capable of anticipating data issues and prepare for them
Data Stewardship Data stewardship encompasses the tactical management and oversight of the company s data assets Data stewardship is generally a business function facilitating the collaboration between business and IT, and driving the correction of data issues
What s Metadata It s more than just data about data Metadata is structured information that describes, explains, locates, or otherwise makes it easier to retrieve, use, or manage an information resource. NISO National Information Standards Organization
Improving Data Quality Management Data Governance Maturity Requirements Maturity++ - Business Glossary - Business Rules - Data Models (Conceptual, Logical, Physical) - Data Dictionary Metadata - Data Lineage - Interface Information - Data Transformations - Data Security Rules - Standards & Frameworks
What s Important to Data Stewards Policies & Procedures Business Requirements and Definitions Data Fitness for Use Data Stewards
Improving Data Stewardship Data Stewardship Data Governance Data Stewardship Data Quality Maturity++ - Business Glossary - Business Rules - Data Models (Conceptual, Logical, Physical) - Data Dictionary Metadata - Data Lineage - Interface Information - Data Transformations - Data Security Rules - Standards & Frameworks
Metadata Management and Data Governance - Context - Business Definitions - Business Rules - Data Quality Rules - Expected Values - Rules and Regulations - Data Usage by Business Processes Metadata Metadata Management Data Governance Business Units Metadata management artifacts are sure to increase knowledge, which is critical to better governance decisions. But the not so obvious is the following: - Ownership Management - Impact Analysis - Audit Trail - Data Lifecycle Management
Reference Data Management Reference Data Management Reference Data Lookup Lists Credit Reports Business Profile Vehicle Catalogs Individual Demographics Auto-Online Auctions Household Information Auto-Physical Auctions Address Reference Auto Used Car Pricing Tax Information OEM Data <more> Governance Integration Architecture & CRUD Sync & Design DQM
CMMI Institute Data Management Maturity Model
FINAL THOUGHTS
Continuous Improvement
Where to Find Me www.datagapconsulting.com www.mdm-in-practice.com www.dcervo.com dalton.cervo@datagapconsulting.com http://www.linkedin.com/in/dcervo @dcervo
THANK YOU