BI: How Can Your High- Performance BI System Meet Expectations When You Feed It 85 Octane Data Making Data Quality Part of the Data Life Cycle Ray McGlew [email protected]
Jun 21, 2015
BI: How Can Your High-Performance BI System Meet Expectations When You
Feed It 85 Octane Data
Making Data Quality Part of the Data Life CycleRay McGlew [email protected]
Brought to You By:
BI Failure Reasons
Gartner: 70%-80% BI Projects Fail• Lack of Business Support and Ownership• Poor Quality Data• Lack of Requirements• Scope Creep• Funding• Big-Bang Approach
Life Cycle of Data
• Creation (usually transactions)• Operational Use• Analytical Use• Destruction
Who “Owns” The Data?
• IT responsible for conserving it– Restrict use according to rules– Providing access– Keeping it safe
• Business responsible for managing it– Create rules for IT to use– Providing IT with requirements for access
• Bottom line… it is a Corporate Resource
Data Concerns
• Privacy – Credit Cards– Health Records
• Security• Accuracy• Usability• Availability
Regulatory Compliance
• Privacy regulations• Legal limits on how long you can keep certain
data• Providing lineage on data used for reporting– Sarbanes Oxley– SEC filings
Quality Data
• Easiest to clean at the source• Some methods to “clean” data– Standardize– Validate
• Data Cleansing Tools
Data Cleansing
Data Governance
“Data governance (DG) refers to the overall management of the availability, usability, integrity, and security of the data employed in an enterprise. “
Data Governance Facets
• Data Stewardship• Data Dictionary/Glossary• Master Data Management• Strong Management Promotion
Data Stewardship
• Splits responsibility for ensuring great data• Business– Defines what important data elements are– Defines the rules for acquiring data– Looks for cross-organizational uses
• IT– Responsible for technical methods– Acquires and maintains tools
Data Dictionary
• Platform for spreading the knowledge• Is used in conjunction with reporting tools• The more data knowledge is used, the better it
gets• Can be started using in-house tools• Starting point for Master Data Management
Master Data Management
• Data Governance should drive MDM• Technology – Facilitates– NOT the driver
Strong Management Promotion
• Cross-functional at the highest level of the organization
• Will require funding• Must break through “It will cost my
department to improve the data quality so their department can save time “
Data Governance and Lifecycle
• Data Creation– Standard values– Validation at the source
• Operational use– Required for some customers and vendors
• Analytical use– Easier to integrate across systems and groups
• Destruction
Data Governance is NOT a Program!
• Culture Change• Integrated with other activities– Business Intelligence– Business Process Re-engineering– ERP Implementation– Mergers
Data Governance Tips
• Prioritize– Based on business value– Based on Pain– Low Hanging Fruit
• Don’t try to boil the ocean!
Resources
DAMA International (www.dama.org)Enterprise Data World
DAMA Philadelphia (www.dama-phila.org)Data Governance (www.datagovernance.com)Data Governance Professionals Org (www.dgpo.org)
Love your data, and stay the course, for it will be with you long after flashy apps are gone.