The Chief Data Officer Management Handbook: Set Up and Run an Organization’s Data Supply Chain

Contents
About the Author
Acknowledgments
Foreword
Preface
Introduction
Part I: Designing an Effective Data Office
	Chapter 1: Understand Your Organization
		Five implicit Data Governance models
			Centralistic Data Governance
			Democratic Data Governance
			Liberal Data Governance
			Technocratic Data Governance
			Anarchistic Data Governance
		Behavioral patterns in data matters
			“Data is an IT task”
			“We can focus on Analytics”
			“It’s digitalization”
			Paralysis by analysis
			“Digital Natives know how to do it”
			“Our business functions can do data on their own”
			“It is all good”
			“Tidy up and tick the box”
	Chapter 2: Aspects of Effective Data Management
		Maturity assessment
		The two main gaps
		Subsidiarity
		Business orientation
		Commercial orientation
		Collaboration
			The Data Office
			Clear data ownership
			A decision and escalation process
			Information sharing
			A Data Stewardship Network
			Checklist
		Motivation
		Cross-functionality
			Focus on organization-wide targets, not on departmental targets
			Incentivize collaboration
			Make the focus part of your organization culture
		Change Management
			Data Literacy
			Help employees understand data
			Share knowledge
			Share data
	Chapter 3: The Data Supply Chain
		A. Manage data sources
		B. Validate data on entry
		C. Classify data
		D. Manage data quality
		E. Do data housekeeping
		F. Curate data
			The purpose of data curation
			Aspects of good data curation
			Provision of information to users
		G. Use data
		Summary: Cover the entire data supply chain
	Chapter 4: Data Vision, Mission, and Strategy
		Data strategy – seriously?
			Strategy vs. execution
			Strategy in times of Agile
			Culture eats strategy for breakfast?
		Vision
			What should a vision accomplish?
			What should a vision focus on?
		Mission
			Define centrally governed data handling standards
			Introduce cross-functional MDM, based on a single source of truth
			Ensure good Data Quality through measurement and improvement initiatives
			Work with business people to turn data into information
			Have all of this supported by the right toolset
			Implement adequate ethical standards in dealing with data
			Train and connect all entities in data matters
		Strategy
			Why do you need a data strategy?
			How is the strategy positioned in comparison to the strategy of the organization?
			How do you develop and maintain your data strategy?
		Your individual measure of success
	Chapter 5: Masterdata Management
		Isn’t static data old-fashioned?
		What does Masterdata cover?
			Masterdata, Reference Data, Metadata
			Examples of Masterdata
			Examples of Reference Data
			Examples of Metadata
		Managing Masterdata
			Cross-functional MDM
			The Data Model
			History view of Masterdata
		MDM and Masterdata software
			Masterdata design styles
			Understand your requirements first
			Determining your requirements
			Waterfall or Agile?
			Build or buy?
	Chapter 6: Data Governance
		Shape a set of Data Principles
		Develop data policies
			What are data policies good for?
			How individual do data policies need to be?
			How do you determine responsibilities for policies?
			How do you determine the setup of your policies?
			How do you develop a set of data policies?
			What does a data policy look like?
		The target state of managed data
		Scope of Data Governance
			Can data be too confidential to be governed?
			Shouldn’t we exempt research data?
			Do we need different Governance for different types of data?
			How about data we don’t know yet?
		Decision-making and collaboration
			Management Board
			Executive data decision body
			Data collaboration group
			Data Community
			Data review and decision process
			Speed of implementation
	Chapter 7: The Data Language
		Characteristics of language
			Don’t we all speak English?
			The dynamics of language
		The data glossary
			What is a glossary?
			The risk of not having a glossary
			What needs to go into a glossary?
			How do you introduce a glossary?
		Data Rules and Standards
			The purpose of rules and standards
			Data Standards
			Data Rules
			Working on Data Rules and Standards
			Documenting Data Rules and Standards
		The Data Model
			The value of a Data Model
			The value of ONE Data Model
			Example customer data
			Analytics and Data Modeling
			Conclusions
		Choosing a software solution
			Do you need a tool to manage data language?
			Are there any primary requirements?
	Chapter 8: Data Processes
		Why prescribing processes?
		Process development aspects
			Ownership
			Existing data processes
			Collaboration
		General considerations
			Technical debt handling
			Conflict management
			Ease of use
			Process triggers
				Request for project approval or funding
				Issue report
				Change request
				Expiry of technical debt
				Request for clarification
				Escalation
			The data review gate
		Concrete process groups
			Data request process
			Project data review process
				When is a project relevant for data review?
				Cases where Data Management is not the primary point of contact
				A typical project data review
			Support processes
			Data content change process
			Data Quality Management process
			Data logic change process
			Data glossary process
			Data access request process
		Manage data in business processes
	Chapter 9: Roles and Responsibilities
		Introduction to roles and responsibilities
		Data Owners and Data Champions
			Data Owners
			Data Champions
		Data Creators and Consumers
		Other business roles
			Business ownership roles
			Lack of ownership
		Centralized roles
			Data Governance
			Data Quality
			Data solutions and projects
			Masterdata management
			Data Architecture
			Data privacy and compliance
			Data Science
			Data Analytics and BI
	Chapter 10: Data Quality
		Why is Data Quality important?
		Dangerous Data Quality standpoints
			Assuming your DQ is good
			Assuming your DQ is good enough
			Assuming you cannot measure DQ
			Assuming bad DQ is a “Data Office” task
			Assuming everybody wants good DQ
			Addressing DQ only once you are in trouble
			Working on DQ for Analytics purposes only
			Working on DQ where first problems arise
			Accepting bad DQ because better DQ is impossible or difficult
			Not communicating the level of DQ
		How to deal with Data Quality?
			DQ must be a top management topic
			DQ requires the right motivation
			Let the right parties raise their hands
			Focus on relevant data
			Keep and get data clean
			Everybody should be responsible
			DQ needs to get measured
			DQ needs to lead to action
		Management of business metrics
			Measure the performance of teams
			Measure the consistency of data
			Consider heuristics
			Determine unwanted behavior
			Break down your quality measurement
			Quality management to cover data
	Chapter 11: Shaping Data Office Teams
		The effective creation of data teams
		Data Architecture and glossary
			The “data language” team
			Glossary Management
			How to organize Data Architecture?
			How to make Data Architecture attractive?
		Analytics
			Analytics across silos
			Data Science
			Data report management
		Document Management
			A centralized approach can add value
			Documents in Data Management?
			How to shape a Document Management team?
		Data Quality
			The central Data Quality team
			Data Quality across the organization
		Organizing Masterdata Management
			Masterdata maintenance
			Masterdata design
			Masterdata coordination
		Data Project Office
			Yet another overhead function?
			Responsibilities of a Data Project Office
			Focus areas
			The Data Project Office within the organization
			Setting up a Data Project Office
		Data service function
			Business helpdesk
			Data organization contact
		Attracting and retaining experts
			Diversity is beneficial – as an outcome
			Everybody wants to join Google
			The sweeter challenge next door
			No meetings, please
			Where’s the infrastructure?
			Playground vs. strategy
			Detached from business
			Little, stupid jobs
			Even Data Science can be boring
			Recognition?
			Data Scientist vs. DB Admin
			Curing the world’s hunger
		Six Sigma
			Six Sigma and data
			Setup of Six Sigma within the Data Office
			A typical DMAIC data process
				Define
				Measure
				Analyze
				Improve
				Control
Part II: The Psychology of Data Management
	Chapter 12: Typical Challenges of a CDO
		Why is it so hard to be a CDO?
		Struggle for supremacy
		Lack of awareness
			WHY should we manage data?
			Why should we manage data NOW?
		Business silos
			Variant 1: “We know best what’s good for us.”
			Variant 2: “I am faster if I do not need to align with others.”
		Lack of ownership
		Opt-out attitude
		Disengagement
		Skepticism
		Business arrogance
		Summary: Prerequisites for success
			Board sponsorship (active!)
			An adequate reporting line
			Clear expectations
			Clear roles in data matters
	Chapter 13: How (Not) to Behave As a CDO
		Don’t rely on formal authority
		Start small, and pick your battles
		Be humble
		Present yourself as a facilitator
		Avoid suboptimal language
		Go out and talk to people
	Chapter 14: Stakeholders
		Manage stakeholders at all levels
			Document your insights
			Classify your stakeholders
			Determine your executive allies
			Know the motives of your allies
			Concrete recommendations
		Tailor your stories to your stakeholders
			Ask the right questions
			Pick the right weaknesses
			Keep data on the agenda
		Shape your data network
			Functional Data Champions
			Business Data Owners
			Data Creators: Data Stewardship Network
			Data Consumers: Analytics Network
			Double loyalty
		Orchestrate your data network
		Plan to consider different audiences
		Frequently stated concerns
			Which problem are you trying to solve?
			What’s in it for me?
			I have no bandwidth for data stuff
			We can look at your strategic ideas tomorrow
			IT has always covered this
			Correct data handling jeopardizes my project
			What if you fail?
			It has worked well without a Data Office
			I don’t want to change…
			Will an algorithm replace me
	Chapter 15: Psychology of Governance
		Don’t claim covered ground
		Design an acceptable starting setup
		Base your authority on accepted authorities
		Balancing two extremes
			Between absolutism and democracy
			Between centralized and local solutions
			Between standardized and individualized
			Between dirty and perfect
		Shape your data brand
		Elevator pitch
Part III: Practical Aspects of Data Management
	Chapter 16: Data Business Cases
		Business cases for data – why?
		Business cases in a perfect world
			The fundamental idea behind a business case
			Capital cost over time
			Consideration of “risk”
			Project selection
			All good?
		General challenges
			Business case culture
			Quantification of benefits
			Conflicting targets within the organization
			Difficult validation in retrospect
			Quickly outdating business cases
		Data-specific challenges
			Data is not sexy
			Determining the benefits of enablers
			Late break-even
			Motives of business leaders
		Eight secrets of data business cases
			I. Do active stakeholder management
			II. Foster data literacy and transparency
			III. Create and maintain a data road map
			IV. Treat data as an asset
			V. Work in stealth mode if necessary
			VI. Explain the cost of NOT doing it
			VII. Set up and follow business case rules
			VIII. Develop a corporate data culture
		Use cases for data as an asset
			What drives your forecasts?
			Data – why NOW
			Consumer data
	Chapter 17: Data Ethics and Compliance
		Ethical behavior and data?
			What could go wrong?
			Where do we stand today?
			Resulting questions for each business
			What are your options?
			A broader perspective
		GDPR – All done?
			You are not done after closing “the project”
			Privacy must become a way of thinking
			See business opportunities
			Recommendations
	Chapter 18: The Outside World
		Why look beyond my organization?
		Sharing data across organizations
			Different ways of sharing data
			The motivation for sharing data
		External data
			Internal data is not enough
			What you can learn from external data
		The CDM and external data
			Challenges around external data structures
			Consequences for your data model
			The mapping of internal and external data
		Data Quality as a service?
		Global standards
			Your own standards – good but not good enough
			Standards across organizations
			Beware of pseudo-standards!
			Practical approach
		Cloud strategy for data
			Outsourcing to the Cloud
			Software as a Service
			Intelligence as a Service
			Risk of Data Model issues
			Risk of black box issues
			Options for proper IaaS usage
		Blockchain
			Unique identification and blockchain
			A false sense of security
	Chapter 19: Handling Data
		The Virtual Single Source of Truth
			What does a VSSoT look like?
			How do you shape a VSSoT?
		Single source of logic
			Service-Oriented Architecture (SOA)
			Distributed logic – the Octopus Principle
		Configuration vs. standardization
		“Effective Date” concept
		Making data international
			The Babylon effect
			Alias terms
			Transliteration
			Country-specific languages
		Data Debt Management
		Agile and data
			A solid foundation – why?
			A solid foundation – how?
		Starting with the happy flow?
			What makes an initiative successful?
			The three dimensions of success
			A scoping approach
	Chapter 20: Analyzing Data
		Preconditions of meaningful Analytics
			Are all preconditions fulfilled?
			Can the investigator influence the outcome?
			How about combining both challenges
		General limits of AI
			Data sources
			AI algorithms
			Human behavior
			AI – Quo Vadis?
		Recommendations around Analytics
			I. Determine the necessary degree of preciseness
			II. Don’t use a formula just because “it works”
			III. Check all preconditions
			IV. Be open about the limitations
			V. Explain your assumptions
			VI. Don’t convey a false impression of preciseness
			VII. Automate data preparation carefully
			VIII. Use DataOps
			IX. Balance diligently
			X. Exclude emotional factors
			XI. Consider changes outside the model
			XII. Define success comprehensively
		Explainable AI (XAI)
			Unknown cause and effect
			Trust issues
			Ethical issues
			Is there a way out?
	Chapter 21: Data Management in Crises
		Prepare for the crisis
			Think the unimaginable
			Be ready to prioritize activities
			Be part of the organization’s crisis plan
		Master the crisis
			Align with company priorities
			Don’t try to become a hero
			Prepare your team
			Listen to your team
			Structure your action
			Manage the state of emergency
		Learn from the crisis
			When does a crisis end?
			The crisis as a catalyst
			Lessons learned
			Celebrate
	Chapter 22: Data in Mergers and Acquisitions
		What is going wrong today?
		Integration planning
		The data approach
			Who should manage data integration?
			Understand the motives
			Focus on interoperability
			Create a high-level plan
			Determine the “best-of-breed” solutions
			Don’t innovate (too much) in parallel
		Data mapping
			Organization
			Concrete mapping cases
	Chapter 23: Data for Innovation
		How can data drive innovation?
			Demystifying innovation
			What is “data-driven innovation”?
			Using data to innovate
		Supporting data-driven innovation
			Determining roadblocks
			Organize Innovation
			Proper handling of business cases
			Adding data to your culture of innovation
		Commercializing data ideas
			The “hundred thousand customers” strategy
			Data innovation factory
Appendix A:
List of Theorems
Bibliography
Index