Master Data Management: Strategic Unity for Your Information Systems Mark Raabe Director of Application Services Faegre & Benson LLP Shawn Samuel Chief Architect Hubbard One Thomson Reuters
Agenda The problem: disconnected information systems How Faegre backed into a solution Advantages of Master Data Management Making the case and getting started Where do we want to take this? Hubbard One: Master Data Mapping Q&A 2
Faegre & Benson Overview 460+ lawyers / 1100+ users 6 offices worldwide: Practice Areas: Corporate Litigation Intellectual Property Healthcare / Nonprofits Finance / Restructuring Employment Real Estate Environmental / Regulatory 3
The Problem is that this is your legal application portfolio. 4
The Problem: Why is it so difficult to get my various enterprise applications to share information? How do I ensure that all systems provide consistent answers? Why are so many of my business units managing duplicative lists? How can I deliver new information systems more quickly? How do we get up and running FAST with new clients, matters, personnel? 5
The Goal: Achieving strategic unity of information systems = Systems that share: Facts Business concepts Rules Objectives = Systems that don t argue with each other. = Systems that enable each other. using the systems I already own. 6
Data integration at Faegre In the mid-1990s, we relied on 1. Staged files in vendor-specified formats 2. Email notifications Manual updates 7
Our first data integration challenge 8
Data integration v1.0 Time & Billing Clients Matters DMS 9
Data integration v1.0 Time & Billing Clients Matters DMS Column Value mdesc1 Hennepin County Arbitration re: Jacob Smith v. mdesc2 ACA International, Inc., a/k/a Audiovisual mdesc3 Corporation of America mdesc1 CLIENT REF # 1996-40-1112 mdesc2 USTM: STAR TRIBUNE - NEWSPAPER OF THE TWIN mdesc3 CITIES 10
Data integration v1.0 Time & Billing Clients Matters DMS Some lessons: Vendors don t know how to design databases! Applications don t care about uses beyond the application Custodians (e.g., Accounting Departments) don t care either 11
Transform Data integration v1.1 Time & Billing clname1 mdesc2 (etc). Clients Matters DMS Column Value MatterName Hennepin County Arbitration re: Jacob Smith v. ACA International, Inc., a/k/a Audiovisual Corporation of America MatterName USTM: Star Tribune Newspaper of the Twin Cities 12
Transform Data integration v1.1 Time & Billing DMS Transform CRM Docket Time Entry 13
The problem in real life: 14
Transform Data integration v1.1 Time & Billing DMS Transform CRM Docket Time Entry 15
Transform Data integration v2.0 Time & Billing Hub DMS CRM Docket Time Entry 16
Transform Data integration v2.0 Time & Billing Hub What have we accomplished so far? Eliminated complexity Standardized matter representation Eliminated dependence on Elite s design Insulated ourselves from Accounting s business practices Reduced load on the Elite server and one more thing: 17
Data integration v2.0 Time & Billing Hub Elite Matter = A way to compile and present the bill Hub Matter = A way to organize all of the Firm s work: Documents / Events / Contacts / Financials / etc. What have we accomplished so far? We have established a master concept of Matter that matches its growing significance to the Firm We have found a way to represent it We have found a place to put it 18
Data integration v3.0 Time & Billing 1a Hub 2 Administrative PPM system 1b How do we improve this concept? 1c 3 1. Protect the data Establish rules: (Data Governance) a. Enforce standardized formats (as set by the standard transform) b. Control who (if anyone) gets to edit data in the Hub directly c. Other systems may consult only the Hub (the System of Record), not Elite directly (the System of Entry) 2. Identify other sources that can contribute matters Elite doesn t have 3. Eliminate alternative matter lists in other applications 19
Data integration v3.0 The industry calls this Master Data Management (MDM): The technology, tools, and processes required to create and maintain consistent and accurate lists of master data. 20
What else goes in the Hub? Well, what else (besides Matters) is Master Data? Transactions = Sentences On 8/25/2010, Attorney X opened Matter Y for Client Z. On 7/7/2010, Paralegal A recorded 3.2 hours to Matter B for this purpose: [narrative] On 2/3/2010, we paid Vendor B $10,000. 21
What else goes in the Hub? then your master data are the shared nouns in your transactional sentences: Matters Hub Vendors? Clients Attorneys Authors Timekeepers People Office Locations Departments & Practice Groups Invoices? Documents? 22
What goes in the Hub? For each master data entity, store the attributes (columns) that are likely to be shared: Matters Matter Name Client Billing Attorney Matter Status Open Date Hub Matters Fee arrangement Bill format Billing address Currency Time & Billing 23
Data integration v3.0 The Master Data Hub : The central repository holding the official version of all of this key information cleaned, de-duped, available for all systems to use and stored in a data structure that describes your business. 24
What else can you do? Aggregate entities that no one source application contains: HR System Attorneys and Staff Hub Summer Associates Recruiting Contractors People Contracting 25
What else can you do? Minimize the pain when an application changes: Time & Billing Hub DMS CRM Docket Time Entry
What else can you do? Find the right source of truth for a master entity: HR System Staff Hub Timekeepers Time & Billing 27
What else can you do? Provision master entities more quickly: New Matter Intake Time & Billing Hub Matters DMS, etc. New Matter Approval Hub Matters DMS, etc. Add Billing Details Time & Billing 28
What else can you do? Identify new master data entities: Attorney adds to My Matters DMS Favorite Matters Hub Favorite Matters CRM 29
What else can you do? Identify new master data entities: Attorney adds to Frequent Matters DMS Favorite Matters Hub Favorite Matters CRM 30
The Business Case for MDM MDM and a Master Data Hub allow you to: Provide master data to all of your applications that is Consistent / Standardized / Accurate Protected Properly normalized Aggregated from all relevant sources Early! Eliminate dependencies on particular applications and vendor data designs Avoid pain, lock-in, and delays! Simplify upgrades, replacements, mergers Understand your own data architecture it s simpler! 31
Transform Transform Perform a maturity assessment Level 0 Level 1 Level 2 Level 3 No (or haphazard)integrat ion between systems Whatever each vendor supplies Point-to-point integration Many paths A Hub containing some master data entities Their meaning is not defined by any single system A strategic Hub that is the basis of the firm s data architecture A comprehensive set of master data entities Time & Billing DMS Time & Billing Hub DMS Governance: Transform CRM CRM Data formats and data quality Docket Time Entry Docket Time Entry System of record a true master Rights to edit
How to start Proceed incrementally (but have a larger design in mind). Pick a target entity. (Clients / Matters / People) Find your existing data flows. How many can be replaced? Define roles. (System of Entry, System of Record, Data Steward, etc.) Design a table for the entity in the Hub. What columns do your consuming systems need? Develop the new data flows. 33
What you need SQL skills / Data modeling skills Data integration tools ETL tools like DTS, SSIS EAI tools like IntApp, BizTalk An eye for patterns and process improvement Political & change management skills Time & patience 34
Some obvious questions: Is there an easier way? Why do I have to build this myself? Won t we all end up with similar designs?
Where should we take this? We need to define standards. The legal industry needs a unified, standard master data model and an API to interface with it. At a minimum, vendors need to recognize and design for MDM: Simplify data interfaces to and from their applications Provide web services where possible Interface with a Master Data Hub, if present Ensure that we always have access to our data The Legal Master Data Hub should be a product. 36
Master Data Mapping and Integration
Master Data Mapping and Integration
Master Data Mapping and Integration
Q & A Keep us posted on your projects! mraabe@faegre.com shawn.samuel@thomsonreuters.com