Incremental Grass Roots Data Governance as an Alternative to Top- Down Directives Andrew Pilch PM/Architect, Founder Pilch Consulting Group, LLC. andrew@pilchcg.com +1 610-984-3296 www.pilchcg.com 1
Andrew Pilch Background Data Governance PM/Architect at Bayer Healthcare MDM Project Manager/Architect at Bayer HealthCare MDM Architect at Merck MDM Architect at Pfizer MDM Business Analyst/Architect at Alliance Consulting Programmer / Analyst at Berlex Laboratories Master's Degree in Information Management from Stevens Institute of Technology Specialties: MDM & CDI Architecture and Project Delivery; Data Governance; Pharma Data & Application Integration; Pharma Sales & Marketing Data Specialist; 2
What is Data Governance? Data Governance is the set of processes that formally manage data assets to ensure the data can be trusted and people are held accountable across an enterprise The factors within a business that affect Data Governance include People, Technology, and Process People Do you trust your employees with your company s data? Technology Does your company have the right technology in place to manage important data assets? Process Does your business operate to properly manage data across the enterprise or does the business operate in silos? Data Governance brings people, technology and process together 3
Pains from Lack of Data Governance Spend and Re-spend Significant investments for initial customer master and warehouse having to be rebuilt or replaced due to turnover and insufficient documentation Missed SLAs Physicians from SFA sent to MDM hub without addresses because manual dependent process not completely run, which results in manual effort to fix and causes downstream processing delays. Mismanagement of Money An undetected vendor data error sent erroneous accrual records, which led to millions of dollars under-budgeting for accruals for years, which resulted in reporting errors and incentive compensation pay-out discrepancies Lack of Data Governance costs money and time 4
Agenda Understand the various types of Data Governance Models Understand why Top-Down Directives for Data Governance fail Understand the Incremental Grass- Roots approach to implementing Data Governance 5
Agenda Understand the various types of Data Governance Models Understand why Top-Down Directives for Data Governance fail Understand the Incremental Grass- Roots approach to implementing Data Governance 6
Data Governance Models (from DGI) Top-Down: Executive Level sets direction for the Enterprise Lower levels of organization follow and add detail Bottom-Up: Lower levels of organization determine policies as part of normal everyday work Details and approval go up the organization hierarchy Center-Out: Dedicated group of experts (e.g. consultants) determine best direction for the Enterprise Executive Level mandates Group works with all levels of organization to implement Silo-In: Representatives from multiple internal groups (Steering Committee) work to set direction and add details Approval is collective and finalized at the Executive Level 7
Agenda Understand the various types of Data Governance Models Understand why Top-Down Directives for Data Governance fail Understand the Incremental Grass- Roots approach to implementing Data Governance 8
Why Top-Down Directives Fail Most executives, directors, and managers are not familiar enough with Data Governance Many organizations do not have the funding or staff to create the optimal formal Data Governance structure across the enterprise: Executive Data Governance Council Strategic Data Governance Steering Committee Tactical Data Governance Group (Data Stewards, System Owners, Subject Matter Experts) Enterprise-wide initiatives take longer to see results because collaboration must occur across many lines and levels of business that may not operate in the same manner 9
Agenda Understand the various types of Data Governance Models Understand why Top-Down Directives for Data Governance fail Understand the Incremental Grass- Roots approach to implementing Data Governance 10
What is Incremental Grass Roots? Grass Roots Data Governance implies a certain set of evangelists within an organization have previously attempted data management policies and procedures within their own group Incremental Grass Roots is to build upon the actions of these evangelists to create data management standards, policies, and procedures within a certain group as a starting point and incrementally add to these as more groups can join, thus eventually becoming the Enterprise model (similar to the silo-in approach shown earlier) Find a good starting point and build from there 11
How do I get started? The first decision to be made using this approach is to pick a good starting group or department. The starting group would have the following basic characteristics: Has professionals that understand the data and have possibly attempted to address issues in the past. (Which group has evangelists?) Has been hurt in the past by bad data quality (Which group s bottom-line is hurting the most due to bad data that can be used as a good business case for Executive sponsorship?) Has Executives that would be willing to sponsor the governance initiative (Pointing out the negative impact to their bottom-line would help get the initiative started, showing improvements to the bottom-line would keep it funded) 12
Group picked, what s next? Once a starting group has been chosen, the next step in this approach is to set up the basic structure : Data Governance Steering Committee (which consists of Business and IT Subject Matter Experts from the initial group) Data Governance Executive Council (which would be the Executive Sponsors of the initial group). The first action for the Committee & Council would be education. In order to be effective as a group, each member should have the same level of understanding of the different aspects of Data Governance, including: Maturity Models Where Are We Going? Frameworks How Do We Get There? Vendors Who Can Help Us Do This Right? 13
Bring in an Expert to Lead the Way In order to have the members of the Council & Committee marching together, you need a leader, a Data Governance expert. Knowing most organizations do not have one, the best choice is to hire an experienced industry expert. The industry expert would make sure all members have an adequate understanding of the various aspects of Data Governance to be effective. The industry expert would also make sure the initiative is moving in the right direction with industry best practices. Do not try this on your own 14
Next Steps, Lots of Discovery The goal of Data Governance is to formally manage your company s important data assets. In order to create policies, processes, and metrics to monitor and report on, you need to first understand the following: What data entities are managed for the group chosen and which are in scope to start? How are these entities defined? Which systems/feeds are associated the data entities in scope and how? What business processes are managed within the group chosen and how do they relate to the systems and data entities in scope? Who are the owners from the Business and IT for each system/data feed in scope? What factors determine success or failure in terms of data flowing from source to consuming systems? Are there existing data quality checks in place to insure success? What are the risks associated with failure? You Need to Know What You Have to Create Policies On 15
Beyond Discovery: Create Standards for Communication of Information During the Discovery Phase, the Council would define information models for the entities in scope (e.g. What is an Individual Business Partner entity, What attributes does a Individual have? ) Having information models sets standards that help map similar attributes named differently across systems 16
Beyond Discovery: Choose a Maturity Model A Data Governance Maturity Model would describe the high-level journey from where your company is currently to where it needs to be in regards to the management of enterprise data assets. A Maturity Model can be used to measure progress as well as create awareness and market Data Governance across the Enterprise Some of the industry best Maturity Models have been developed by the following groups : MDM Institute Data Flux Gartner IBM Choose the Model that Best Fits Your Organization 17
Beyond Discovery: Choose a Framework A Framework would be used to guide the group on the details of implementing of Data Governance; (Who, What, When, and Where) Frameworks assist in describing and organizing complex concepts and their interrelationships Some of the industry best Frameworks have been developed by the following groups : Data Management Association (DAMA) Data Governance Institute (DGI) IBM Frameworks work with a Maturity Model to describe the breath and depth of a Data Governance initiative Choose the Framework that Best Fits Your Organization 18
Bottom Line: Start Small and Build Up Find a good starting group/dept: Has attempted data management in the past Has been hit financially by bad data management Has the capability and willingness to put funds and people in place to make improvements Get help! Bring in an industry expert to guide the way Make use of industry-leading methods for implementing Data Governance (maturity models & frameworks) Maintain and Evolve: Create awareness across the enterprise Build upon initial group s structure, standards and policies with the next group Make Data Governance a Reality Not a Dream 19
How to Leverage PCG? Provide guidance through all phases of the Data Governance initiative Identify and train Council & Committee Members as well as Data Stewards Perform system/data analysis for Discovery as well as the creation of metadata Guide the architecture of implementing Data Governance policies Andrew Pilch PM/Architect, Founder Pilch Consulting Group, LLC. andrew@pilchcg.com +1 610-984-3296 www.pilchcg.com Our Focus is Data Quality 20
Q&A Questions? Comments? 21