Data for Business Performance: The Goal-Question-Metric (GQM) Model to Transform Business Data into an Enterprise Asset, by Prashanth Southekal
Master how to leverage your data to improve business performance.
Understanding the GQM Model
Understanding Data
Understanding the Enterprise
Associating Enterprise and Data
Understanding the Asset
Understanding Transformation
Data and Business Value
Decision Making
Compliance
Customer Service
What has Changed?
Conclusion
Business Data in the Enterprise Value Chain
Primary activities
Support activities
Business Data versus Non-business Data
Types of Enterprise Business Data
Reference data
Master data
Transactional data
Metadata
Key Characteristics of Business Data
One Data, Many Views, Many Classes
Purpose
Origination
Processing
Analysis
Time horizon
Sensitivity
Ownership
Treatment
Lifecycle
Security
Conclusion
Business Data Lifecycle
Origination
Capture
Validation
Processing
Distribution
Aggregation
Interpretation
Consumption
IT Functions
Storage
Security
IT Systems
Databases
Relational or SQL databases
Non-relational or NoSQL databases
SQL or NoSQL?
OLTP Systems
Integration Systems
EAI (Enterprise Application Integration)
ETL (Extraction, Transformation, and Loading)
OLAP Systems (Business Intelligence Systems)
Analytics Systems
Relating OLAP and OLTP Systems
Managing the Systems
SoR (Systems of Record)
SoD (Systems of Differentiation)
SoI (Systems of Innovation)
The Big Picture
System Architecture — The Data Model
Conclusion
Key Limitations of Data
Data is normally obscured and biased
Data doesn’t always translate into actions and results
Relevancy of data is a function of time, space, and stakeholder type
Data has the potential to cause “analysis paralysis”
Stakeholders’ needs precede metadata and data dictionary ontology
Data management is expensive and time-consuming
Data might distort innovation
Data is never real-time; it is always historical
Data has no relevance for first-time events
Data can mislead decision making
Conclusion
Data Quality Dimensions
Completeness
Consistency
Conformity or validity
Uniqueness or cardinality
Accuracy
Correctness
Accessibility
Data security
Currency and timeliness
Redundancy
Coverage
Integrity
Consequences of Poor Data Quality
Data Depreciation and its Factors
Causes of Poor Data Quality
Data silos resulting from organization silos
Interpretation and consumption of data happen in different ways
Frequency of use and the number of users
Poor business case for data origination and capture
Data searching and retrieval challenges
System proliferation and integration issues
Different value propositions between consumers and originators of data
Data rules affect business operations
Data quality is time-sensitive
Results of data quality improvements are normally transient
Data conversion and migration issues
Interface feeds
System upgrades
Manual errors
Poor database design
Data purging and cleansing
Conclusion
Deep Dive of the GQM Framework
Conceptual level — the goal
Operational level — the question
Quantitative level — the metric
Leveraging GQM for Data Management
Step 1: Conduct stakeholder analysis
Step 2: Formulate the stakeholders’ goal(s)
Step 3: Identify the parameters in the goal statement
Step 4: Translate the goals and parameters to quantifiable questions
Step 5: Answer the questions and derive the hypothesis
Step 6: Derive attributes from the questions, answers, and hypothesis
Step 7: Derive metrics from the attributes
Step 8: Apply the reverse GQM or MQG framework (Metric-Question-Goal)
Step 9: Derive the data elements from metrics or answers
Step 10: Profile the data elements and build the data model
A Simple GQM Example
Case Study 1: Application of the GQM for Insights
Step 1: Conduct stakeholder analysis
Step 2: Formulate the stakeholders goal(s)
Step 3: Identify the parameters in the goal statement
Step 4: Translate the goals and parameters to quantifiable questions
Step 5: Answer the questions objectively
Step 6: Derive hypotheses and attributes
Step 7: Derive metrics from the attributes
Step 8: Apply MQG framework
Step 9: Define data elements from the metrics
Step 10: Profile the data elements and build the data model
Case Study 2: Application of the GQM for Compliance to Industry Standards
Step 1: Conduct stakeholder analysis
Step 2: Formulate the stakeholder goal(s)
Step 3: Derive the parameters for the goals
Step 4: Translate the goals to quantifiable questions
Step 5: Answer the questions
Step 6: Derive hypothesis and attributes
Step 7: Derive metrics from the attributes
Step 8: Apply the MQG framework
Step 9: Derive data elements for these metrics
Step 10: Profile the data elements and build the data model
Conclusion
Reference Architecture
Principles
Principle 1: Data is managed for a purpose
Principle 2: Quality business data is an enterprise asset
Principle 3: Realizing quality data takes investment
Principle 4: Enterprise data has clear ownership
Principle 5: Data shall always be accessible and shared in an enterprise
Principle 6: Enterprise data is secure
Patterns
Baseline the current level of data management maturity
Data-driven initiatives should be tied to a strong business case and business KPIs
Enterprise goals should precede LoB goals
Manage core business processes in the SoR
Reference and master data should be based on data standards and MDM
Data integration (EAI and ETL) should be specific to the data types and the business rules
Distribute reporting in OLTP, BI, and analytics systems
Data security practices should be integral to business operations
Enterprise data governance (EDG) should be an active and functional business entity
Data-driven culture should be enterprise-wide
Standards for Data Management
Chief Data Officer (CDO)
Conclusion
Complying with Regulations Using the COBIT Model
Compliance with Industry Standards
Compliance to Internal Policies
Business Insights and the DIKW Model
Intuition and Decision-Making
Enterprise Analytics
Small Data, Big Data and Hadoop
Transforming Data into Insight
Identify the problem domain and define the goal
Formulate questions and hypotheses
Derive KPIs
Identify and profile data elements
Collect and normalize the data
Visualize the data
Analyze the dataset
Synthesize the dataset
Interpret and validate insights
Communicate insights (data storytelling)
Analytics in the Real World
Conclusion
Stakeholders in the Data Lifecycle
Managing Change
Aligning LoB leadership to the Enterprise
Managing Change in Individuals and Teams
Understanding the key deliverables of the stakeholders
Identifying the elements that pertain to the behavior of the individual
Understanding the change process
Transitioning to the desired state
Conclusion
Today, digitization is dramatically changing the business landscape, and many progressive organizations have started to treat data as a valuable business asset. While many enterprises are investing in improved data management, only a few have leveraged data to truly impact business performance. To address this problem, Data for Business Performance provides readers with practical guidance and proven techniques to derive value from data in today’s business environment. Specifically,
this book:
This book is absolutely timely and relevant in today’s data-driven world. Most of the books on data available in the market today focus on data quality, governance, and analytics. This book from Dr. Prashanth Southekal is brilliant as it puts the business stakeholder at the center by addressing the key value propositions of the business user. This book is holistic and I strongly believe it will help to bridge the gaps we have today.
Mario Faria
Managing Vice President, Gartner, US
Packed with insights and leveraging a process oriented approach, this book covers a unique combination of the science, the art and the strategy of unlocking the potential of data for enterprises in a real-life context. The author has managed to provide a clear action plan for creating data analytics and its management a key function in a modern enterprise.
Ashish Sonal (Vir Chakra)
CEO, Orkash, India
This book is one of the most practical sources for how companies can greatly improve their bottom line by improved data management and becoming a data-centric company. It combines leading data management theory with step-by-step implementation and real-life examples, and is a must-read for those wanting to derive more value from their corporate data.
Lance Calleberg
Application Architect, Husky Energy, Canada
Prashanth has given a very practical guide to implement data culture in an organization. The book Data for Business Performance talks about building the organization of the future and the role of data. Prashanth rightly believes and demonstrates that data is not an asset of the IT team and is an organization-wide asset. He proposes the need for the chief data officer (CDO) as a role that should anchor data and report to the CEO, and manage the stakeholders’ data needs.
Harshajith Umapathy
Senior Vice President, Hansa Cequity, India
Dr. Southekal provides valuable insights on data and information management in mostly short and clearly written sections. Anyone interested in the data-driven company should read this book and learn about the hurdles on the road to be data-driven, and his valuable suggestions on how to overcome them. His wisdom may prevent some of the failures that helped him learn.
Erik van der Voorden
Domain Architect, Independent Consultant, Netherlands
Please complete all fields.