Master Data Management (MDM) Primer and Best Practices by Mani Kumar Manda & Eugene Breger 26 th September, 2011 Mastering Master Data Solutions 1 Agenda for the Today s Presentation This presentation covers following topics: 1. MDM Primer Session 1 2. MDM Best Practices Session 2 Mastering Master Data Solutions 2
Introduction Mastering Master Data Solutions 3 About Mani Kumar Manda Mani Kumar Manda Rhapsody Technologies, Inc. Naperville, IL, USA Founder and Chair for OAUG CDM SIG http://cdmsig.oaug.org http://groups.yahoo.com/group/cdmsig President of Rhapsody Technologies, an Oracle partner with specialization in implementing Oracle s MDM applications Frequent speaker of topics associated with Oracle s MDM products Customer Data Hub, Product Hub, Site Hub and Supplier Hub Working with Oracle Applications for over 17 years and has implemented Technology Solutions for clients in many industries. Mastering Master Data Solutions 4
About Eugene Breger Sales Director for Rhapsody Technologies, an Oracle Partner with specialization in implementing Oracle s MDM applications Eugene Breger Rhapsody Technologies, Inc. Naperville, IL, USA Working with Oracle Applications for over 18 years and sold technology-based business solutions to clients in many industries Chicago-based Oracle employee at both Oracle and PeopleSoft leading Consumer Packaged Goods and Retail Business Unit initiatives Mastering Master Data Solutions 5 About Rhapsody Founded in 1998 Managed by a team of veterans with over a century of collective experience A player in MDM space even before the term MDM was coined Preferred Partner of Oracle in MDM space Implementation Services Implement Oracle R12 Applications Oracle MDM Hubs Customer Product Supplier Site COA Oracle Fusion Applications Packaged and Assessment Services Packaged Services Oracle DQ Discovery Oracle DQ Search Optimization Oracle TCA Functional Support Build Semantic Business Model for Customer Assessment Services Master Data Health Check MDM Readiness Assessment MDM Implementation Health Check Training Services MDM Boot Camp 1 or 2 Days format Addressing Master Data Quality in R12 EBS implementations Half Day & Full Day Executive MDM Seminars Half Day & Full Day Oracle Customer Data Management Functional Fundamentals Oracle Customer Data Stewardship Training Customers (Partial List) Mastering Master Data Solutions 6
Session 1 of 2 MDM Primer Mastering Master Data Solutions 7 Agenda for Session 1: MDM Primer 1. Master Data 2. Master Data Management (MDM) 3. Integration Landscape 4. Defining 360 0 view 5. More about MDM 6. Key Concepts in implementing MDM 7. Architectural Styles 8. Conclusion Mastering Master Data Solutions 8
Master Data Mastering Master Data Solutions 9 Data is Structured Unstructured Data Data Mastering Master Data Solutions 10
Structured Data Hierarchy Meta Data (Data about Data) Structured Data Audit Data Archive Data Rhapsody Reference Data Examples: Lookup Codes Transactional Data Examples: Orders, Invoices, Leads, Master Data Examples: Customer, Product, Analytical Data Examples: Facts, Cubes, Dimensions, Mastering Master Data Solutions 11 Then Reference Data has longest life span changes rarely May change due to external factors o Country List (break up USSR for example) May change due to internal factors o Change in Invoice Types due to new way of doing business for example usually low volumes Tens of rows to hundreds, rarely in thousands requires little on-going effort to maintain can be one of following four (4) types Codes that are external to the enterprise such as Country Codes Codes that are used as classification schemes such as industry classifications, market segmentation Lookup Codes such as Status code, Type code, Invoice Type Constant values, could be external or internal o External : Tax Rates o Internal : Employee optional deductions Mastering Master Data Solutions 12
Then Master Data has a long life span changes gradually over time both due to internal and external factors Customer: Name changes, Customer moves, phone number changes, contact changes, etc. Product: New Version due to ECO, additional attribution, etc. Supplier: Name changes, Supplier moves, phone number changes, contact changes, etc. requires an on-going effort (requires programs, not just projects) Need significant effort in people, process and technology to keep Master Data at optimal quality requires Business leadership to keep it at optimal quality requires IT leadership to provide technology and tools to business to maintain it on an ongoing basis Mastering Master Data Solutions 13 Mastering Master Data Solutions 14
Master Data Management (MDM) Mastering Master Data Solutions 15 Per Rhapsody, MDM means About establishing the Single Source of Truth for Master Data, in other words a System of Record (or Reference) that is reliable, available when needed and is always maintained About consolidating, de-duping, and cross-referencing the Master Data across heterogeneous systems to establish a Single View of the Master Data About providing 360 0 view of the Master data About establishing a Data Governance framework for welfare of the Master Data solution by establishing Roles and Responsibilities; Policies and Procedures to steward the master data on an ongoing basis About facilitating the definition and maintenance of Analytics and Segmentation to drive growth in top line (Revenue) and bottom line (Profits) by increasing the effectiveness of various business processes Increasing the operational excellence across the Enterprise Accomplish all of the above with a synergistic alliance between Business and IT Mastering Master Data Solutions 16
Master Data Management (MDM) is According to The CDI Institute The authoritative, reliable foundation for data used across many applications & constituencies with the goal to provide a single view of the truth no matter where it lies. According to Gartner Master Data: Consistent and Uniform set of identifiers and attributes that describe the core entities of the enterprise and are used across multiple business processes MDM is a technology-enabled discipline in which business and IT organizations work together to ensure uniformity, accuracy, stewardship, semantic consistency and accountability of the enterprise s official, shared master data assets. According to Forrester Master data is the most trusted and unique version of important enterprise data, such as customer, product, employee, or asset that is captured, maintained, and used across disparate systems. MDM is the business capability that delivers master data throughout an organization through the alignment of multiple information management technologies such as PIM, CDI, and data quality, plus business process improvements and organization commitments. Mastering Master Data Solutions 17 MDM Categories & Domains (or Entities) Parties Things Places Concepts Customer Supplier Employee Citizen Partner Bank Product Part SKU Service Asset Facility Location Geography Chart of Account Contract* License Policy Previously known as Customer Data Integration (CDI) Vendors sell as Customer Hubs and Supplier Hubs aka Product Information Management (PIM) Vendors sell as Product Hubs Vendor sell as Site Hubs (Only Oracle has a pre-built application) Solution for Chart of Accounts is most popular Mastering Master Data Solutions 18
MDM Domains Key Assets of an Enterprise Forms the foundation for Operational and Analytical (DW & BI) Systems Are enterprise-wide Cross-Business Unit, Cross-functional, Cross-Departmental Mastering Master Data Solutions 19 According to Gartner 1, MDM Initiatives in Customer Domain is HOT in Discrete Manufacturing Life Sciences Retail Financial Services Health Care Government Product Domain (Sell Side) is HOT in Discrete Manufacturing Process Manufacturing Life Sciences Communication Financial Services 1 As of Q1 of 2010 based on presentation given by Gartner at 2010 Gartner MDM Summit in Las Vegas, Nevada. Mastering Master Data Solutions 20
Integration Landscape with and without use of MDM Hub Application Mastering Master Data Solutions 21 Integration Landscape without MDM Legacy Mktg. Sales EIS Quoting DW OM Service AR Contracts Collections Credit Mastering Master Data Solutions 22
Integration Landscape with MDM Hub & Spoke Model Legacy Mktg. Sales EIS Quoting DW MDM Hub(s) OM Service AR Contracts Collections Credit Mastering Master Data Solutions 23 Mastering Master Data Solutions 24
Defining 360 0 view Mastering Master Data Solutions 25 What is 360 0 view of the Master Data? Rather than thinking about the Customer itself, we need to think about the relationship that exists with the Customer, and that lets us provide Superior experience. Mastering Master Data Solutions 26
What is 360 0 view of the Customer? Quotes Orders Invoices Receipts Service Contracts Installbase Agreements Etc. Identity Attributes Name, Address, Education, etc. Behavioral Attributes Hobbies Tech Savvy, Old Timer, etc. Financial Attributes Income, Net worth, etc. Classification Attributes Industry, Size, etc. Relationship Attributes Account Number, Terms, etc. Segmentation Attributes Categorization, etc. Corporate Hierarchies Business specific hierarchies Past relationships Staff and Family Information Affiliation with Professional Associations Mastering Master Data Solutions 27 Short Video about Challenges of Customer Data Mastering Master Data Solutions 28
Corporate Asset One of the most valuable asset of any Organization is its customer information. Everyday millions of dollars in potential profits are lost at many organizations due to inability to provide reliable, accurate, well maintained and readily available customer data. Mastering Master Data Solutions 29 Mastering Master Data Solutions 30
More about MDM Mastering Master Data Solutions 31 Drivers for MDM Business Side Top-line Growth (Increase Revenue) Bottom-line Growth (Increased Profit) Superior Customer Experience Enhanced Customer Loyalty IT/Enterprise Objectives Regulatory Compliance Privacy Management Risk Management Cost Savings Quality Data Mastering Master Data Solutions 32
Benefits of MDM Hard Benefits Obtain a consistent and accurate master data See a 360 0 view of the master data when applicable Avenue for better regulatory compliance Better management of Privacy Laws Facilitates the quicker assimilation of acquired companies in M&A scenario Mastering Master Data Solutions 33 Benefits of MDM Soft Benefits Increased efficiency of business processes o Results in cost savings Improved Customer Satisfaction o Results in higher customer retention Greater Share of Wallet o Results in increased revenue Perception of Company as Customer Centric o Results in higher customer retention Better Target Marketing o Results in higher return on marketing spend Mastering Master Data Solutions 34
MDM Solutions Vendor Approaches Application Centric Philosophy: Each MDM Domain is unique and requires an application specifically developed for it Key Proponents o Oracle o SAP Platform Centric Philosophy: All MDM domains have some commonalities and a user needs a technology platform using which they can create and manage any MDM Domain Key proponents o IBM o Tibco o Informatica (through Siperian) o Hyperion DRM An exception: Oracle Hyperion DRM (A Platform Centric, but Oracle views it has a strong contender for certain MDM Domains as well as complimentary technology for existing MDM Hubs of Oracle Mastering Master Data Solutions 35 Key Questions Products and Pricing How can I tell which products a customer is buying across all my businesses? Can I put together groups of products that customers will want to buy? How can I differentiate my pricing by client segment? Marketing & Sales How are clients contacting my business and what are they buying through each channel? How do I expand the type of products that my existing customers buy? How do I find out which customers to sell more products to and which to ignore? Operations & Servicing How can I minimize current data quality and consistency issues in the company? How do I keep customers satisfied enough to keep buying our products? What is the most effective way for the company to communicate with a customer? How can I avoid asking customers for the same information that they ve already provided to the company? How do I understand the correct level of service to provide to my customer? How can I tell if a customer has other holdings within the company that impact his credit score? Technology How should I manage the replacement of legacy information systems in the next five years? How can I simplify the process of combining our information with that of another business? How can I control the sprawling cost of managing enterprise data? Regulatory Compliance Lack of system synchronization is causing data quality issues, which is preventing me from implementing effective process controls to comply with SOX and other regulations? How do I address this? Mastering Master Data Solutions 36
Key Concepts Mastering Master Data Solutions 37 Terminology Basic Terms System Generated Data is derived by the system processes based on other data. Ideally done in One System System of Entry The system within which a given data topic or segment or attribute is captured via a CRUD operation (Create, Update, Delete). Can be done in Multiple Systems System of Record A system serving as the authoritative source for a given data topic or segment or attribute. Only One System System of Reference The recommended system serving as a source of data within a given context for a given data topic or segment or attribute among several possible sources of data. Should be One System, Reality is many Mastering Master Data Solutions 38
Terminology Basic Terms System Generated Data is derived by the system processes based on other data. Ideally done in One System System of Entry The system within which a given data topic or segment or attribute is captured via a CRUD operation (Create, Update, Delete). Can be done in Multiple Systems System of Record System of Reference A system serving as the authoritative source for a given System of data Record topic also or segment serves as or System attribute. of Reference In order for a System to be considered as System of Reference, this system should be (a) a System of Record The or (b) recommended the data should system have serving come as directly a source from of the data within System a given that context is considered for a given System data topic of Record. or segment or attribute among several possible sources of data. Only One System Should be One System, Reality is many Mastering Master Data Solutions 39 Terminology Federated & Centralized Federated System of Entry is done in Spoke Applications and the data is synchronized with Hub and other Spokes. Changes can be made in Spokes and Hub Applications subject to System of Record agreements. DW Legacy Credit Mktg. Hub AR Sales OM Quoting Mktg. Legacy Sales Centralized System of Entry is done in the Hub and the data is synchronized with Spokes. DW Hub Quoting Spokes are not allowed to make changes. Credit OM AR Mastering Master Data Solutions 40
Architectural Styles for Master Data Mastering Master Data Solutions 41 Architectural Styles - Overview Master Data Point to Point SOA Hub without SOA Without Hub With Hub Hub in Operational App Federated Centralized Confederated Regional & Enterprise Hubs Stand Alone Hubs Hubs in Operational Apps Mastering Master Data Solutions 42
Mastering Master Data Solutions 43 Conclusion Mastering Master Data Solutions 44
To Conclude Master Data Management = Data Governance + Data Consolidation + Data Quality + Profiling + Cleansing + De-Duping + Metrics + Data Enrichment + Data Integration Mastering Master Data Solutions 45 To Conclude The impact of MDM projects will be felt across the Enterprise and should be treated as an Enterprise level initiative MDM initiatives takes long time to get buy-in as well as implement. Don t expect to rollout an MDM solution quickly Learn from the experience of others and plan accordingly Fixing data quality challenges is time consuming and often times feels like a daunting task. Make sure to set expectations accordingly Don t underestimate the political currents and roadblocks that gets thrown frequently Do not buy MDM products first without getting into deeper analysis of pain points and requirements Most importantly choose an MDM technology for all MDM domains of importance to your organization, even if the initial implementation is for a specific domain or domains Mastering Master Data Solutions 46
To Conclude Mastering Master Data Solutions 47 Crystal Ball (5 to 10 years down the line) The value of data will be treated as an Asset on the balance sheet (Accounting Wise) Data Governance will become mandatory (Regulatory perspective) Reporting on data quality metrics becomes a necessary KPI in financial statements CXO s held responsible for data quality lax in quality could become a punishable offense CGO Creation of a Chief Governance Officer as an executive level position with CGO possibly reporting to CEO CPO Chief Privacy Officer possibly reporting to CGO This role exists in some web-centric organizations IT will be looked upon as a key enabler in meeting above requirements Mastering Master Data Solutions 48
Q U E S T I O N S A N S W E R S Mastering Master Data Solutions 49 B R E A K Mastering Master Data Solutions 50
Session 2 of 2 MDM Best Practices Mastering Master Data Solutions 51 Agenda for Session 1: MDM Primer 1. Prologue (Before implementation) 2. During Implementation 3. Epilogue (After first implementation) Mastering Master Data Solutions 52
Best Practices Prologue Mastering Master Data Solutions 53 Best Practices Prologue 1. Making a Business Case 2. Stakeholders with Enforcement Authority 3. Obtain Executive (CXO) Sponsorship Mastering Master Data Solutions 54
1. Making the Business Case Data Profiling Regulatory Landscape Analytical Use Cases Create Upside potential vs. avoid downside Risks Mastering Master Data Solutions 55 1. Business Case - Data Profiling Choose a strategically important application for the Master Data Domain of interest Conduct Data profiling of the master data in this application Use these metrics to identify the impact on the Organization Invest in a third party tool to facilitate the data profiling which will be used extensively before, during and after implementation of any MDM solution Trillium (a unit of Harte-Hanks) Oracle Warehouse Builder (OWB) Talend (Open Source) Mastering Master Data Solutions 56
Before Profiling The perception.. Mastering Master Data Solutions 57 After Profiling - The shocking revelation Mastering Master Data Solutions 58
1. Business Case - Regulatory Landscape Identify the regulatory landscape that dictates the need to manage master data Look into Privacy Laws o Customer Child Labor Laws o Supplier Environmental Laws o Site o Supplier o Product OSHA Regulations o Site Denied Party List (DPL) Mandate o Customer o Supplier Mastering Master Data Solutions 59 1. Business Case Regulatory Landscape - DPL Compliance Denied Party List (aka DPL) is a list of persons and entities with whom no one should be doing business with In US Department of Treasury o Specially Designated Nationals and Blocked Persons (SDN) o Narcotics Trafficking Sanctions o Anti-Terrorism Sanctions o Non-proliferation Sanctions Bureau of Industry and Security o Denied Persons List o The Entities List FBI o Most wanted list o Issued Watch List US Department of State o Debarred Parties List o Designated Terrorist Organizations Mastering Master Data Solutions 60
1. Business Case Regulatory Landscape - DPL Compliance Mastering Master Data Solutions 61 1. Business Case Regulatory Landscape - DPL Compliance Mastering Master Data Solutions 62
1. Business Case - Analytical Use Cases It is important to take into consideration of requirements (in other words use cases) of Business Intelligence and Data warehousing applications, which are downstream applications to MDM application Could be as simple as data that is basis for Segmentation May Influence other analytical requirements Mastering Master Data Solutions 63 1. Business Case Upside Potential vs. Downside Risks Business Case must include Use Case scenarios for Upside Potential of instituting and implementing an MDM program o Data Profiling will provide some scenarios Use Case scenarios for Downside Risks of inaction o Study of Regulatory Landscape will provide scenarios Mastering Master Data Solutions 64
2. Stakeholders with Enforcement Authority Establish Data Governance Council Data Governance Council must have stakeholders that have enforcement authority Otherwise the good decisions that are made by the council get lost resulting in project failures You should include Departmental Heads, Business Unit Leaders, at least one of key CXO Governance council is not the place for some one who does not have enforcement responsibilities. Senior persons with correct background who want to take part or need to take part in MDM initiative that do not have enforcement responsibility can be part of Data Steward Team. Mastering Master Data Solutions 65 3. Obtain Executive (CXO) Sponsorship To succeed you must have CXO (executive) support for all MDM initiatives Many programs and projects that do not have CXO support or lost support soon after launching the initiative have failed An increasing number of initiatives in MDM are facing challenges (Reference Gartner: The Top Seven Trends for Master Data Management in 1Q08) Not only Securing but ongoing support of Executives is very important for long term success Making a strong Business Case upfront and socializing it across the organization will help you not only to secure but to sustain the Executive Sponsorship Mastering Master Data Solutions 66
he who does not lay his foundation beforehand, may by dint of great ability lay them afterwards, yet it must be with the most extreme labour on the part of the architect and with the greatest danger to the building taken from The Portfolio - Vol. 7 (1819) Mastering Master Data Solutions 67 MDM Best Practices - Prologue Q & A Q U E S T I O N S A N S W E R S Mastering Master Data Solutions 68
Best Practices During Implementation Mastering Master Data Solutions 69 Best Practices Implementation 1. Launch Data Governance Initiative Strong Data Stewardship 2. Hybrid Project Implementation Methodology 3. Analysis Semantic Model Conceptual Data Model Data Security Data Profiling 4. Data Cleansing Data Standardization Validate Addresses 5. Design Implementation Approach Data Quality Metrics Design Considerations Canonical Designs Classify, Classify, Classify!!! Robust Search UI Go beyond the Identity Data, in stages 6. Conversion Enrich Data Mastering Master Data Solutions 70
1. Implementation - Launch Data Governance Program Establishing Data Governance is all about establishing the framework for the decision rights and accountability for the welfare of the data Data Governance initiative is a must for any MDM project(s) You must establish the Data Governance Program at the beginning of the project Data Governance program should be designed at the Enterprise level addressing multiple domains of MDM, not just for Customer or Product Both Business and IT must come together Base framework established by the time Analysis phase is done, but no later than early design phase Mastering Master Data Solutions 71 Data Governance and Data Management Framework and Structure CEO EXECUTIVE EMPOWERMENT Sales Marketing Service Impact Analysis IT CXO(s) Rhapsody Finance HR Membership Data Security ERP & CRM HR & Other EXECUTIVE Council Rhapsody Data Governance Council Change Control Compliance Rhapsody Data Governance Liaison Data Stewards Data Management Data Quality Experts Functional SME s App. Experts BI/Analytics SME s Project Management Cust. Data Experts Documentation Training Data Management Organization Rhapsody Mastering Master Data Solutions 72
Data Governance Business and IT Come Together Business IT Mastering Master Data Solutions 73 1. Implementation - Strong Data Stewardship Data Stewardship team is the one facilitating many decisions associated with data elements related to a specific MDM data domain What attributes to host in MDM instance? What entities should be brought over to MDM? What is the phasing approach? To what extent standards should be enforced for Source Applications Departmental Politics Identify Data Stewards internally Provide training to internally recruited Data Stewards Mastering Master Data Solutions 74
2. Implementation - Hybrid Project Implementation Methodology Many MDM projects often take years to complete to obtain the vision of 360 0 view at an Enterprise level, if at all one reaches that destination Risk of losing executive and/or user support Approach the project with global design but phased implementation Mature the MDM initiative in an incremental manner Mastering Master Data Solutions 75 Project Implementation Approach Spiral Methodology Support Evaluate Lessons Learned (LL) Maintain 4 3 Support & Evaluate 2 1 1 2 4 Requirements & Analysis 3 Incorporate LL Requirements Validation & Prioritization As-Is Study Acceptance Test Plan 1 1 2 2 Build/Develop, Unit Test Integration Test System Test, UAT, Training Production Rollout 3 Implement 4 3 Design 4 Architecture Design CRP1, 2, Mastering Master Data Solutions 76
3. Implementation - Analysis - Semantic Business Model A semantic business model is a knowledge base developed with concepts and the relationships between these concepts based on how business is conducted Example: A semantic business model for Customer/Supplier data can be built from a business perspective with an objective to identify the people and the entities and all relationships between them. This should be done at an early stage of the MDM project. For Customer/Supplier domains, this model can be called Trading Community Model (TCM) or Semantic Customer/Supplier Model The Semantic Models are not based on any application rather they are about business o The TCM will facilitate the decision of what people, entities and relationships to host in MDM for Customer Instance Surprise Factor o At several clients, this task revealed surprises: Good portion of people, entities and relationships were never captured in any application in the first place. Of those missing, some are very critical to the business No single person or department has complete view of the Trading community. It often took interviewing about 100 to 200 people from various business units and departments to be able to put this portrait (picture) together Mastering Master Data Solutions 77 A sample Semantic Customer Model of a Financial Institution Reachable via Nominates Represents Promotes thru Promotes thru Current/Former Customer of Collaborates With Services the Sells/Services To Sells/Services To Sells/Services To Sells/Services To Influences to buy Mastering Master Data Solutions 78
3. Implementation Analysis - Conceptual Data Model The flexible functionality of an MDM solutions leads into the ability to create same set of data in multiple ways For example in Oracle Customer Data Hub (CDH), the TCA framework allows you to create customer data in many ways If you take a customer with 3 locations o Option#1 Site Centric Model Create One party with 3 locations o Option#2 Party Centric Model Create Three parties with each party having one location and relate these parties to form a (corporate) hierarchy o Option#3 Some where between Option #1 and Option #2 Customer Model Usability Mastering Master Data Solutions 80 Option #1 - Site Centric Model Widely known as Old AR model Release 11 Model Mastering Master Data Solutions 81
Option #2 - Party Centric Model Party 2 Party 1 Account 2 Party 3 Account 1 Site 2 Party Site 2 Account 3 Site 1 Party Site 1 Site 3 Party Site 3 Mastering Master Data Solutions 82 3. Implementation Analysis - Data Security Data security is important for a global MDM solution Carefully evaluate users in terms of their access requirements Utilize built in security functions of the tool For example Data Sharing and Security (DSS) functionality in Oracle CDH Role based security in PDH, SDH, Site Hub, etc. Make sure to identify all regulatory requirements for data Security Doing otherwise will lead the firm into regulatory nightmare Recent Examples o Lost data BNY Mellon Shareowner Services lost Tapes containing SSN data o Hacks Sony PlayStation Network Hack costs $170 Million TJ Maxx (TJX) pegs data breach tab at $118 Million Mastering Master Data Solutions 83
3. Implementation Analysis - Data Profiling Conduct comprehensive data profiling in each of the source systems, an extension of the profiling done during business case Data Profiling is a must do task Done prior to Conversion usually during early stages of Analysis Done periodically after go-live to monitor the state of data quality Data Profiling helps in identifying The good data that can be brought over to an MDM Application Bad data that can be ignored or fixed prior to conversion Attributes that can be Standardized Identify the data that is misrepresented Invest in a third party tool to facilitate the data profiling Trillium (a unit of Harte-Hanks) Oracle Warehouse Builder (OWB) Talend (Open Source) Mastering Master Data Solutions 84 4. Implementation Data Cleansing - Data Standardization It is important to standardize the data for better data quality Use third party tools for data standardization for Customer and Supplier Data Use Oracle PDQ Server to standardize Product descriptions Or build custom routines to standardize the data prior to or as part of conversion Not all tools can help you standardize For example, CDH can not standardize the data But can standardize the data for purposes of identifying duplicates only Mastering Master Data Solutions 85
4. Implementation Data Cleansing - Validate Addresses It is critical to validate and standardize the address data to facilitate better duplicate identification Use the Vendor whose solution is pre-integrated with the tool For example o Trillium and First Logic for CDH o Many vendor solutions for UCM through hooks Mastering Master Data Solutions 86 5. Implementation Design - Implementation Approach Depending on vendor solution there may be one or two options available MDM Stand Alone environment Decisions can be made independent of operational system Can apply patches when needed Implementation cycle not dependant upon operational application MDM Operational environment (For example ebusiness Suite-EBS) Reduced Licensing Costs Inability to apply patches when needed Implementation cycle is influenced by operational needs of other modules Modeling impact MDM SOA EBS EBS MDM Mastering Master Data Solutions 87
5. Implementation Design - Data Quality Metrics No MDM initiative will be fully successful, unless Data Quality objectives are set upfront The quality of the initial data is measured The data can be monitored over a period of time to identify the increase in data quality The project must include the time and resources for defining Data Quality Metrics, Procedures that facilitate the tracking of Data Quality Metrics and reporting of improvement in data quality Mastering Master Data Solutions 88 5. Implementation Design - Design Considerations Example: Customer domain Your Customers Business Design Considerations Your Business How you Intend to do Business? Mastering Master Data Solutions 89
5. Implementation Design Canonical Designs Adopt an open standard based Canonical Mapping for integrating data between Source Applications and CDH and CDH and Target Applications Open Applications Group (http://www.openapplications.org) The Canonical format should be accommodating needs at enterprise level, not at departmental or a specific application level Global Design with Phased Implementation If needed, start with a standards based canonical model and customize it further Some products support open standards CDH with their Application Integration Architecture (AIA) and Process Integration Packs (PIP) leverages OAGIS standards Mastering Master Data Solutions 90 5. Implementation Design - Classify, Classify, Classify!!! It is important to identify all classification requirements of all consuming applications Classify data as many ways as important to the business Classify at an appropriate level Party Level classification is not suitable for Account level entities Product Category Classifications vs. Product Catalog Classifications Classification will facilitate the better intelligence of numbers such as Sales, Receivables, etc. Mastering Master Data Solutions 91
5. Implementation Design Robust Search UI Should be able to query based on all key attributes Common perception is to provide key identity attributes You should also provide the ability to search based on Source System References Ability to filter the result set by source systems Role based Search Parameters Mastering Master Data Solutions 92 5. Implementation Design - Go beyond Identity Data In Stages Identity data is a must to achieve the single view of the Customer Even then Bring one source at a time into MDM Data for which an MDM solution for Customer is suitable is Account Layer Industry specific information o Healthcare Hospital information such as specialties, number of beds, physicians, Segmentation Make sure that the MDM instance matures over a period of time in multiple phases Mastering Master Data Solutions 93
Replication of Master Data Usually this data is rarely replicated across applications A good portion of the data is replicated across many applications Most of this data is replicated across applications Mastering Master Data Solutions 94 6. Implementation Conversion - Enrich Data When to do enrichment? During conversion; or Post Conversion Data enrichment is often necessary step for many businesses D&B data to build corporate hierarchies in case of Customer Domain Industry specific content providers Healthcare o Verispan o Solucient o Sherlock Multiple providers may be needed some times Mastering Master Data Solutions 95
MDM Best Practices During Implementation Q & A Q U E S T I O N S A N S W E R S Mastering Master Data Solutions 96 Best Practices Epilogue Mastering Master Data Solutions 97
Best Practices Epilogue 1. Certify Data 2. Latest Versions 3. Socialize Results 4. Ensure Continued Executive Sponsorship Mastering Master Data Solutions 98 1. Epilogue - Certify Data Provide a means to identify the data that has been reviewed and validated Search results should show the certification status of the data Various capabilities in tools CDH allows to mark the data with custom defined certification levels and reasons Not all tools for all data domains have this capability Mastering Master Data Solutions 99
2. Epilogue - Latest Versions The overall MDM is still maturing resulting in change in approaches and functionalities in a product The products in this space are rapidly increasing their functionality as well as fixing bugs Must upgrade to a latest Version Upgrading to R12.1.2 in case of CDH Upgrading to 8.2 in case of UCM Latest versions provide new capabilities o Business Object level integration vs. granular integration in R12 o Governance Manager in case of UCM 8.2 Must also apply all latest patches High number of Bugs requiring frequent patch application Spend time upfront in identifying patches in the functionality that you plan to utilize Mastering Master Data Solutions 100 3. Epilogue Socialize Results Publicize the success of the MDM implementation Promote the operationalization of MDM capabilities in operational and analytical applications The more an MDM application is tapped the more important it becomes to the organization Mastering Master Data Solutions 101
3. Epilogue Ensure Continued Executive Sponsorship It is important to validate and publicize the ROI after implementing an MDM solution Ensure to publicize the increase in upside potential due to an MDM application as well as the mitigation of downside risks again due to an MDM application Promote usage of MDM application to benefit more parts of the organization in turn to secure continued executive support Extend MDM initiatives to new data domains Mastering Master Data Solutions 102 Q U E S T I O N S A N S W E R S Mastering Master Data Solutions 103
Mastering Master Data Solutions 104 Q & A Q U E S T I O N S A N S W E R S Can be reached at: Mani Kumar Manda Email: MManda@RhapTech.com Phone: 630-717-1809 Special Interest Groups: Customer Data Management SIG http://cdmsig.oaug.org http://groups.yahoo.com/group/cdmsig http://groups.yahoo.com/group/ucmsig Eugene Breger Email: EBreger@RhapTech.com Phone: 847-331-5554 Product Data Quality SIG http://www.linkedin.com/groups?mostpopular=&gid=3350438 (http://tinyurl.com/29xocwv) Supplier Life Cycle Management (SLM) & Supplier Data Hub (SDH) SIG http://www.linkedin.com/groups?mostpopular=&gid=3363004 (http://tinyurl.com/2v2yyct) Oracle Fusion Applications http://www.linkedin.com/groups?mostpopular=&gid=2316926 (http://tinyurl.com/3232emw) Mastering Master Data Solutions 108