Data Architecture.

1. Introduction
● Architecture, in a general sense, is an organized arrangement of components intended
to optimize the function, performance, feasibility, cost, and aesthetics of a structure or
system1. The International Organization for Standardization (ISO) defines architecture as
"the fundamental organization of a system, embodied in its components, their
relationships to each other and the environment,2 and the principles governing its design
and evolution"3.
● The goal of Data Architecture is to be a bridge between business strategy and
technology execution4. Data Architecture is considered from three perspectives:
outcomes (artifacts like models and definitions), activities, and behavior (collaborations
and skills)5.
● Data Architecture artifacts include specifications that describe the existing state, define
data requirements, guide data integration, and control data assets according to a data
strategy6. An organization's Data Architecture is described by master design documents
at various levels of abstraction, including standards for how data is collected, stored,
arranged, used, and removed7.

Business Drivers for Data Architecture: 8

● Strategically prepare organizations to evolve products, services, and data for new
business opportunities.
● Translate business needs into data and system requirements.
● Manage complex data and information delivery throughout the enterprise.
● Facilitate alignment between Business and IT.
● Act as agents for change, transformation, and agility.

Data Architecture Outcomes and Practices

● Definition: Data Architecture involves "Identifying the data needs of the enterprise
(regardless of structure), and designing and maintaining the master blueprints to meet
those needs. Using master blueprints to guide data integration, control data assets, and
align data investments with business strategy"9.
● Goals: 10
1. Identify data storage and processing requirements11.
2. Design structures and plans to meet the current and long-term data requirements of
the enterprise12.
3. Strategically prepare organizations to quickly evolve their products, services, and
data to take advantage of business opportunities inherent in emerging

1
 technologies13.
● Practices: Data Architects aim to bring value through an optimal technical footprint,
operational efficiencies, and the organization's increased ability to use its data14. This
involves defining and maintaining specifications that:
○ Define the current state of data15.
○ Provide a standard business vocabulary16.
○ Align Data Architecture with enterprise strategy17.
○ Express strategic data requirements and outline designs to meet them18.
○ Integrate with the overall enterprise architecture roadmap19.

Essential Concepts
Enterprise Architecture Domains
Data Architecture operates alongside other domains, and architects from each must
collaborate20.

Domain Purpose Elements Dependencie Roles

s

Enterprise To identify Business Establishes Business

Business how an models, requirements architects and
Architecture enterprise processes, for the other analysts,
creates value capabilities, domains. business data
for customers services, stewards.
and events,
stakeholders. strategies,
vocabulary.

Enterprise To describe Data models, Manages data Data

Data how data data created and architects and
Architecture should be definitions, required by modelers, data
organized and data mapping business stewards.
managed. specifications, architecture.
data flows,
structured
data APIs.

2
 Enterprise To describe Business Acts on Applications
Applications the structure systems, specified data architects.
Architecture and software according to
functionality of packages, business
applications in databases. requirements.
an enterprise.

Enterprise To describe Technical Hosts and Infrastructure

Technology the physical platforms, executes the architects.
Architecture technology networks, application
needed to security, architecture.
enable integration
systems to tools.
function and
deliver value.

The Zachman Framework for Enterprise Architecture

● The Zachman Framework is a well-known enterprise architectural framework developed
by John A. Zachman25. It is an ontology represented by a 6x6 matrix that describes the
complete set of models required to describe an enterprise26.
● The framework's two dimensions are:
1. Communication Interrogatives (Columns): What (inventory), How (process),
Where (distribution), Who (responsibility), When (timing), and Why (motivation)27.
2. Reification Transformations (Rows): These represent the perspectives of different
stakeholders: planner, owner, designer, builder, implementer, and user28.
● Each cell in the matrix represents a unique design artifact created from a specific
perspective to answer one of the fundamental questions29292929.

Enterprise Data Architecture (EDA)

● EDA defines standard terms and designs for important organizational data elements30.
Its design includes depicting business data, its collection, storage, integration,
movement, and distribution31.
● EDA descriptions must include two key components:
1. Enterprise Data Model (EDM): An EDM is a "holistic, enterprise-level,
implementation-independent conceptual or logical data model providing a common
consistent view of data across the enterprise"32. It includes key data entities,
relationships, business rules, and some attributes33. Successful EDMs are often built

3
 incrementally and iteratively in layers34.
2. Data Flow Design: Defines the requirements and blueprint for data storage and
processing across databases, applications, and platforms35. Data flows illustrate how
data moves through business processes and systems from its origin to where it is
stored and used36. They can be represented in matrices or diagrams37.

Activities
Data architecture addresses complexity from two viewpoints: a quality-oriented focus on
improving execution within development cycles and an innovation-oriented focus on
business and IT transformation38383838.

● Establish Data Architecture Practice: This involves selecting frameworks, developing a

roadmap, organizing the work, defining best practices, and producing artifacts39. Key
steps include:
○ Evaluate Existing Data Architecture Specifications: Identify and evaluate existing
documentation for accuracy and completeness40.
○ Develop a Roadmap: A roadmap provides a way to manage data dependencies and
make forward-looking decisions, describing the architecture's 3-5 year development
path41414141.
○ Manage Enterprise Requirements within Projects: Development projects
implement data solutions based on business requirements and EDA standards42. This
involves defining project scope, understanding requirements, designing solutions,
and implementing them, whether by buying, reusing, or
building434343434343434343434343434343434343434343.
● Integrate with Enterprise Architecture: Data Architecture priorities are often driven by
funded projects44. It is crucial to integrate EDA matters with project portfolio
management and enterprise application development planning45454545.

Tools
● Data Modeling Tools: Necessary for managing the enterprise data model, these tools
often include functions for tracking lineage and relationships between models46.
● Asset Management Software: Used to inventory IT systems and track their
relationships, this software contains valuable metadata for creating data flows and
researching the current state47474747.
● Graphical Design Applications: These are used to create architectural diagrams, data
flows, and other artifacts48.

Techniques

4
● Lifecycle Projections: Architectural designs should be clearly documented to represent
their status, such as Current, Deployment Period, Strategic Period, or Retirement49.
● Diagramming Clarity: To be effective, diagrams must use visual conventions
consistently50. This includes using a clear and consistent legend, matching objects to the
legend, maintaining clear line direction, minimizing line crossings, and ensuring linear
symmetry515151515151515151.

Implementation Guidelines
● Implementing EDA involves organizing teams, producing initial artifacts (e.g., enterprise
data models, data flows), establishing working methods in projects, and creating
awareness of the value of these efforts52.
● Readiness and Risk Assessment: Architecture initiation projects carry significant risks,
including:
○ Lack of management support53.
○ No proven record of accomplishment54.
○ An apprehensive sponsor who micromanages communication55.
○ Counter-productive executive decisions made from a lack of understanding56.
○ Culture shock among employees affected by changes57.
○ An inexperienced project leader58.
○ Dominance of a one-dimensional view (e.g., from owners of a single ERP system)59.
● Organization and Cultural Change: The speed of adoption depends on the
organization's culture60. Success depends on factors like cultural receptivity, recognizing
data as a business asset, adopting an enterprise perspective, and integrating
architectural deliverables into project methodology61.

Data Architecture Governance

● Enterprise Data Architecture and Data Governance organizations must be well-aligned62.
Ideally, a data architect and a Data Steward should be assigned to each subject area63.
● Governance Activities Include:
○ Overseeing projects to ensure compliance with architecture standards64.
○ Managing architectural designs, their lifecycle, and tools65.
○ Defining standards, rules, and guidelines for data use66.
○ Creating artifacts that enable compliance67.
● Metrics: Performance metrics for EDA reflect architectural goals and are often
monitored annually68. They include:
○ Architecture standard compliance rate: Measures how closely projects adhere to
established architectures69.
○ Implementation trends: Tracks improvements in the organization's ability to

5
 implement projects, including the reuse of artifacts and project execution
efficiency70.
○ Business value measurements: Tracks progress toward business benefits, such as
improvements in agility, quality, operational efficiency, and the business
environment71717171.

Acronyms and Definitions

Acronym Meaning / Definition from Chapter

Context

BOM Bill-of-material 72

BSP Business Systems Planning (an IBM

method) 73

COTS Commercial Off The Shelf (purchased

applications) 74

CRM Customer Relationship Management 75

CRUD Create, Read, Update, and Delete data 76

DAMA Data Management Association

International 77

DMBOK Data Management Body of Knowledge 78

EDA Enterprise Data Architecture 79

EDM Enterprise Data Model 80

ERP Enterprise Resource Planning 81

6
IEC International Electrotechnical Commission
82

ISO International Organization for

Standardization 83

ISP Information Systems Planning (a method

popularized by James Martin) 84

PDM Product Data Management 85