Logical Modeling for an Enterprise MDM Initiative Session Code TP01 Presented by: Ian Ahern CEO, Profisee Group Copyright
Speaker Bio Started career in the City of London: Management accountant Finance, Banking & Securities Strong focus on financial consolidation, statutory reporting & risk management systems 25 years of Software & Consulting experience: Data warehousing & Business Intelligence Financial Modeling Master Data Management Focused on MDM for last 15 years Delivered the technology that is now Microsoft s Master Data Services embedded in SQL Server Started Profisee is 2007 to continue pushing forward Master Data solutions 2
Agenda & Topics Introduce Profisee What is Master Data and Master Data Management? The challenges of data modeling in a MDM initiative Building your own MDM Center of Excellence
Profisee Experience & Heritage 10+ years :: Master data software releases 100+ :: Master data implementations worldwide Since 2001 :: Deep customer & industry master data experience 2001 2006 Stratature formed, +EDM delivers multi-domain MDM & hierarchy management Microsoft Gold ISV partnership established 2007 Microsoft acquires Stratature Profisee formed by core Stratature executives 2010 Maestro released alongside SQL Server Master Data Services 2011 Gartner lists Profisee as Cool Vendor in MDM 2012 Maestro Industry Solutions released 2013 Maestro Adapters released Advanced modeling & ERwin integration STRICTLY CONFIDENTIAL
Representative Maestro Customers STRICTLY CONFIDENTIAL
Master Data is Non-transactional data A noun/entity used within the business Used to describe any transaction, group of transactions, or other data entity Slowly changing, relative to the overall rate of transactions Any relevant reference data Any associated attributes, properties, or relationships that define, classify, describe, or enhance master data Master Data Management therefore contains a substantial amount of entity and attribute modeling combined with respective metadata.
Master Data is People Things Places Abstract Customers Customers Products Locations Accounts Vendors Vendors Business Units Stores Warranties Sales People Bill of Materials Wells IP Employees Parts Power Lines Metrics Partners Media Geo Areas Securities Patients Equipment Warehouses Contracts
Master Data Management Measurement recognizes the importance of metrics to sustain focus and progress. This includes data quality measurement and individual performance management considerations. Technology specifies the target MDM architectural style evolution and identifies how existing technologies may be used for the solution. It also elaborates the role of a serviceoriented architecture. Governance Directives that manage the organizational bodies, policies, principles, and qualities to promote access to accurate and certified master data Organization depicts the roles, responsibilities, and relationships of those participating in the program. Standards establish the guidelines and rules for creating and maintaining master data. This includes rule-sets for data architecture, data modeling, and the creation of data domains and elements. Process specifies new processes to be introduced and identifies considerations for modifications to existing master data processes which will be required to achieve the target state.
The challenges of data modeling in a Master Data Management initiative Copyright
The Traditional MDM Application Lifecycle (From the modeling perspective) The simplified view Logical Modeling & Physical Modeling Physical Database Application Server & Clients Too simplified to work in practice
Key MD Modeling Challenge: Adaptive & collaborative modeling through the entire lifecycle It's important to note that traditional approaches to Master Data Management (MDM) will often motivate the creation and maintenance of detailed LDMs, an effort that is rarely justifiable in practice when you consider the total cost of ownership (TCO) when calculating the return on investment (ROI) of those sorts of efforts. Extracts from 12 Critical Lessons in Agile Data Modeling 1. Agile modelers create agile models which are just barely good enough. 2. Agile developers solve today s problem today and trust they can solve tomorrow s problem tomorrow. 3. Agile data modeling is both evolutionary and collaborative. 5. Agile data models can and should follow your corporate standards. 6. Trying to define all the requirements upfront is a risky proposition.. Interesting data modeling challenges. at literally every step in the MDM application lifecycle. Acknowledgement: Scott W. Ambler http://www.agiledata.org/essays/agiledatamodeling.html
The Traditional MDM Application Lifecycle (A push of logical to the physical to the application) The more realistic view Master Data Architect / Data Steward / Implementer Logical & Physical Modeling Database, Services & Apps Data Architect Data Architect DBA this is collaboration, but it is not Agile.
An Adaptive / Agile MDM Application Lifecycle (Data governance processes) The agile requirement Internal MDM Services, Workflows and Processes Logical & Physical Modeling Database, Services & Apps Data Architect Data Architect DBA Master Data Architect / Steward / Implementer External Services ERP/ CRM etc. Many stewards, employees and even external agents, customer or partners interacting with MDM data Subject to role based security: Any user, workflow, system or process should be able to adapt master data. And, if necessary, the underlying model structure and properties.
An Adaptive / Agile MDM Application Lifecycle (Three potential modeling processes) Logical Modeling Physical Modeling Adaptive Modeling Data Architects Master Data Design & Implementation Application Users & Processes ERwin Maestro Modeling Surface Maestro Adaptive Modeling A completely integrated Agile modeling environment.
Industry Models Entities, Metadata & Reference Data Healthcare Oil & Gas Hospitality & Gaming Insurance Retail Financial Services Accelerate implementation & time-to-value Leverage pre-built models & industry standards Integrate existing models & entity relationship diagrams Incorporate industry best-practices for MDM Reduce project risk & total cost of ownership
The Value in Industry Models Patterns, Entities, Relationships Reference Data Entity Model for Diagnosis Diagnosis Classification ICD 10 Great for patterns Metadata is useful in electronic interchange scenarios Excellent if from a highly reliable source Look for mappings between versions of reference data
Points to Consider on Industry Models Balance data, process and use case requirements. Don t let a data model or pattern preference dictate an entire physical application design. Check sources of reference data. (Entity by entity if necessary) Validate any professional memberships or subscriptions required for ongoing access. Obtain the source for any integration to obtain data & populate the tables. At some point you will need to customize this code for a future requirement. Ensure there is an easy way to hand pick & utilize patterns, groups of entities etc. Be careful of defaulting a physical model to an industry model. Generally, an industry model should assist design process not dictate a physical design. Opt for productized application layer rather than instance-specific application coding. Deliver higher return on investment and shorter time-to-value. Easier to maintain an application in the longer term if less pieces break with change.
Building your own MDM Center of Excellence Copyright
The Cornerstone of Information Management Single Version of the Truth Almost every piece of valuable information in an organization is identified, calculated, stored, retrieved, analyzed, reported and utilized based on its categorization by master data. Without a single version of master data, creating a single version of information is virtually impossible. Accuracy Every system in an organization from an ERP to a spreadsheet uses master data in order to store and display information in a meaningful way. Inaccurate or inconsistent master data, or its use across systems, is the largest driver of errors in reports and analytical information. Timeliness Integration Consistency Consolidation Quality Productivity Mapping
Goal Benefits Drivers Master Data Applications Master Data Value Map Maximize Shareholder Value Increase Revenue Decrease Cost Optimize Assets Acquire New Customers Grow Existing Customers Improve Pricing Strategy Reduce Cost of Goods Sold Streamline Processes Optimize Productivity Improve Facility, Inventory & Employee Management Improve Data Efficiency & Reusability Address/Contact Cleansing & Verification Bill of Material Optimization Branch/Store Performance Customer Golden Records/Single View Chart of Accounts Standardization Facility Utilization Customer/Product Rationalization Cost Management Incentive Management Householding & Relationship Analytics Credit Risk Management Inventory Optimization Marketing & Promotion ROI ERP Consolidation/Migration Sales Analytics Merger & Acquisition Management Product Golden Records/Single View Talent/Workforce Management Product Information Management Supplier/Vendor Data Management Territory Planning
Master Data Management Solutions Master Data Management Solutions Common Challenges Ideal Solutions Data Quality Compliance Improve Efficiency Retain Customers M & A Improve Decisions Cross Reference Golden Records Incomplete, inaccurate, duplicate customer data ERP implementation, consolidation, migration Tracking spending by customer State and federal mandates Different types of customer accounts Tracking customer purchase history Identifying customers and relationships Incorrect, outdated customer contact data Merging charts of accounts Acquisition inclusion vs. exclusion analysis Inconsistent, incorrect contact & address info Relationships between customers & parents Conflicting customer data across multiple systems Tracking product sales across channel/territory Distributed/siloed customer/product/supplier data Inaccurate/duplicate customer data in source CRM Data matching, standardization, de-duplication ERP integrated multi-domain MDM & governance Mastered customer/product/sales data Transparent & auditable statutory reporting Single view of all customer data Customer/product/sales data integration & reporting Customer golden records & hierarchies Contact/address verification & standardization Chart of accounts standardization COA standardization, scenario/version management Address/contact data validation & completion Mastered customer>parent hierarchies & mappings Survivorship of validated consolidated customer data Global product & performance consolidation Single view of any master data domain Customer data cleanse/match/survive/harmonize
Think Big Targeted Execution Natural Flow from Enterprise MDM Thinking into Targeted Execution 1 2 3 4 Assessment Roadmap Accelerator Production Launch Guidance for near term strategic and tactical MDM objectives New to MDM, preparing an initiative and wanting to ensure best practices from the outset Delivering Executive and Technical Summaries Detailed surveys of business, systems and technology and 6 disciplines Potential prototype High-level MDM project plan An Enterprise Roadmap and the basis for a Center of Excellence and a multi-year, phased approach to an envisioned end-state Aligning MDM with strategic and tactical goals Plans for the 6 disciplines and associated matrix of governance roles, stakeholders and policies Data quality assessments Define projects and plan Create a self-sufficient center of MDM excellence Design, document, and deliver initial domain via a phased prototype to production approach Complete classroom curriculum for MDS and Maestro technologies Just in time skills transfer, assignments, review, and best practice guides & audits Focused delivery of short-term ROI Joint responsibility for full deployment to production Complete review of hardware & software configuration Review load balancing, network optimization, and data bus integrity Propose, review and audit application lifecycle management for models and data Optimize production Conduct stakeholder reviews post deployment Enterprise MDM Envision & Plan Phase 1 Implement and Deploy
Profisee MDM Professional Services Proven Methodology Fixed-price Packages Engagement Approaches Mentoring & skills transfer Turnkey application Center of Excellence Accelerate model development, integration & deployment Leverage best-practices of over 100 successful MDM projects Optimize solution scalability, performance & usability Attain MDM self-sufficiency via hands-on skills transfer Reduce project risk and internal resource strain
Thank You Questions? Ian Ahern CEO, Profisee Group ian.ahern@profisee.com www.profisee.com