Dgi Framework
Dgi Framework
Dgi Framework
Abstract
Data Governance can mean different things to different people. Adding to this
ambiguity, governance and stewardship can be perceived as complicated endeavors.
Frameworks help us organize how we think and communicate about complicated or
ambiguous concepts. If your organization employs a framework, your people can
more easily achieve clarity of thought and purpose. A framework can also
help you succeed in realizing value from your program and efforts and data.
The DGI Data Governance Framework was designed to help you:
Achieve clarity
Ensure value from your efforts
Create a clear mission
Maintain scope and focus
Establish accountabilities
Define measurable successes
Components
ofused,
a Data
Governance
Program
10: Processes 10
for governing
how data is
and when,
how, and by whom.
Components of
the DGI
Framework
6. Defining
Data Data Governance
10. Stakeholder
Care and
1. Aligning Policies,
Requirements, and Controls
2.People
Establishing Decision Rights
3. Establishing
Accountability
&
4. Performing Stewardship
Organizational
5.Bodies
Managing Change
7. Resolving
Issues
WHO
Support
11. Stakeholder
Data Quality
Data
Data 8. Specifying
7
9
Communications
Requirements
Data
Data
Governance Office
Stakeholders
Stewards
and
9. Building
Governance Into(DGO) 12. Measuring
Governance
Reporting Value
Technology
WHY
Rules
&
Rules of
Engagement
Processes
Mission
WHAT
2 Goals
Focus
Areas
Metrics /
Success
Measures
Funding
to
achieve
WHEN
1
Develop
a value
statement
2
Prepare
a
roadmap
3
Plan
and
Fund
4
Design
the
program
5
Deploy
the
program
6
Govern
the
data
7
Monitor,
Measure,
Report
HOW
Business / IT Processes
that touch data
Decision Rights
Accountabilities
Controls
5
6
10
Management
This paper describes core concepts, the components of the DGI Data Governance
Framework, and typical steps in implementing a program.
Table of Contents
What is Data Governance?.............................................................................................................. 3
Data Governance and IT Governance ......................................................................................... 4
Why Use the DGI Data Governance Framework? ......................................................................... 5
What Do You Want Data Governance to Accomplish? ................................................................. 6
Typical Data Governance Focus Areas....................................................................................... 6
Data Governance With a Focus on Policy, Standards, Strategy ............................................. 7
Data Governance With a Focus on Data Quality .................................................................... 8
Data Governance With a Focus on Privacy / Compliance / Security ..................................... 8
Data Governance With a Focus on Architecture / Integration ................................................ 9
Data Governance With a Focus on Data Warehouses and Business Intelligence (BI) ........... 9
Data Governance With a Focus on Management Support .................................................... 10
Your Focus and Your Stakeholders .......................................................................................... 10
Working Toward Your Goals With a Data Governance Life Cycle Methodology ...................... 11
Framework Components in Detail ................................................................................................ 12
Data Governance Components that Deal With Rules and Rules of Engagement ................ 13
Component #1: Mission and Vision ..................................................................................... 13
Component #2: Goals, Governance Metrics / Success Measures, Funding Strategies ......... 14
Component #3: Data Rules and Definitions ......................................................................... 15
Component #4: Decision Rights ........................................................................................... 15
Component #5: Accountabilities........................................................................................... 16
Component #6: Controls ....................................................................................................... 17
Data Governance Components that Deal With People and Organizational Bodies ................ 17
Component #7: Data Stakeholders........................................................................................ 17
Component #8: A Data Governance Office (DGO) ............................................................. 18
Component #9: Data Stewards.............................................................................................. 18
The Process of Governing Data ................................................................................................ 18
Component #10: Proactive, Reactive, and Ongoing Data Governance Processes ............... 18
Challenges ..................................................................................................................................... 19
Getting Started .............................................................................................................................. 20
Page 2 of 20
Definition
Data
Governance
is the exercise
of decisionmaking and
authority for
data-related
matters.
Whatever you focus your efforts on, youll need to define for
your participants exactly what you mean by Data Governance.
Choose your words carefully; they must resonate in your
culture and environment if your program is to be accepted.
t Does your organization have strict hierarchical,
command-and-controls protocols? Are you operating
in a compliance-driven, black-and-white
environment? Do you know what you want, but the
challenge is getting people to obey existing rules? If so,
your definition may emphasize exercise of authority.
t What if youre still trying to create collections of
policies, rules, and data definitions? What if your issues
fall more into a shades-of-gray area, where the
challenge is getting the right participants to
collaboratively agree on the right set of rules, and to
work together to monitor and enforce them? In that
case, your definition may emphasize decision rights.
Its important to present a definition that implies what you are
trying to accomplish with your program. You dont want to
send the wrong message to those who are helping to govern
your data, or to those who are being governed.
Alternate
Definition
Management
is the decisions
you make.
Governance is
the structure
for making
them.
- CIO Magazine
Page 3 of 20
Soundbite
Governance
comes into
play when
individual
managers find
that they
cannot or
should not
make
independent
decisions.
Soundbite
IT is like the
pipes and
pumps and
storage tanks
in a plumbing
system.
Data is like the
water flowing
through those
pipes.
If you
suspected
your water
was poisoned,
would you call
a plumber?
Page 4 of 20
Goals
Page 5 of 20
Benchmark
The most
common
objective
of Data
Governance
programs
is to
standardize
data
definitions
across an
enterprise.
Perspective
The type of
policies you
address and
the level of
involvement
required of
data
stakeholder
groups will
depend on
the focus of
your Data
Governance
Program.
Page 6 of 20
Soundbite
Formal Data
Governance
policies,
backed by
crossfunctional
Stewards,
can give
needed
weight to
architectural
positions.
Page 7 of 20
Soundbite
Data
Governance
often sets
direction for
Data Quality
and then
monitors the
success of
Data Quality
efforts.
Soundbite
Programs that
focus on
Privacy /
Compliance /
Security often
stem from a
management
mandate.
Page 8 of 20
Soundbite
A move to
ServiceOriented
Architecture
(SOA), with its
need for wellgoverned
data, could
be a driver for
a program
with a focus
on
Architecture /
Integration.
Soundbite
Data
Governance
programs with
a focus on
Data
Warehouses /
BI often start
small, but may
scale to other
efforts
The idea is to
act locally
but think
globally.
Page 9 of 20
Soundbite
Data
Governance
programs with
a focus on
Management
Support
typically
come into
existence
when
managers find
it difficult to
make
routine
data-related
management
decisions
because of
their potential
effect on
operations or
compliance
efforts.
Page 10 of 20
2
Prepare
a
roadmap
3
Plan
and
Fund
4
Design
the
program
5
Deploy
the
program
6
Govern
the
data
7
Monitor,
Measure,
Report
Note that Data Governance does not begin with the design of
the program!
t Before you start deciding who goes on what committee,
you should be clear about your programs value
statement.
t You should have developed a roadmap to share with
stakeholders.
t Those stakeholders will want to know
the WHO / WHAT / WHEN / WHERE / WHY / HOW
of your program before they decide to support it, so you
need to anticipate their questions. Youll need
preliminary answers, even if theyre only assumptions
until you do your actual program design.
These first three phases of your Data Governance Life Cycle
are difficult. Weve all heard the saying that its hard to see the
forest for the trees. What if you ARE one of the trees?
If youre just getting started, consider asking for assistance
from another group within your organization that has
successfully launched a new program. What lessons did they
learn about value statements and funding?
Best Practice
Consider
receiving
input from
other
organizations
that have
successfully
launched
Data
Governance
programs.
Page 11 of 20
Benchmark
Page 12 of 20
Data
Governance
Page 13 of 20
Best Practice
To be
successful in
establishing
and enforcing
policies,
standards,
and other
types of rules,
your Data
Governance
program will
need to
provide the
type of
support thats
built into the
DGI Data
Governance
Framework.
Page 14 of 20
Best Practice
Bring discipline
to your
program
by translating
goals to
formal value
statements.
Page 15 of 20
who will have to live with the decision? Of course not that
decision needs to be made by the organizations Board of
Directors, with input from the Legal department.
For other types of programs, decision rights may require more
negotiation. For example, who should decide the length of a
data field in a new system? Hmmm maybe the decision
should be made by Data Architecture. But maybe it requires
input from many stakeholders. Maybe one of them has a
constraint that needs to drive the decision.
Component #5: Accountabilities
The PostCompliance
Paradigm
This new
paradigm says
that, for efforts
with a
compliance
requirement,
work is not
finished until
you
1) Do it,
2) Control it,
3) Document
it, and
4) Prove
compliance.
Page 16 of 20
Background
Controls can
be:
Preventative
Detective
Corrective.
They can
manual,
technologyaided, or
completely
automated.
Best Practice
Consider
implementing
a Data
Governance
Board made
up of
executivelevel
stakeholders.
Page 17 of 20
Definition
Data
Stewardship:
the set of
activities
that ensure
data-related
work is
performed
according to
policies and
practices as
established
through
governance.
Page 18 of 20
Challenges
At the industrys first Data Governance Conference, in
Orlando, Florida USA in December of 2006, leaders of
successful Data Governance programs declared that, in their
experience, Data Governance is between 80% and 95%
communications! They uniformly expressed surprise at the
amount of written and verbal communication required to
successfully bring together stakeholders to achieve their goals.
Is communication not your fort? Thats OK. Communication
plans and supporting tools can be developed by those with
expertise and then administered by your program staff.
For example, most DGOs employ customer segmentation to
understand stakeholders information needs, preferred
terminology, and special interests. They develop layers of
communication pieces elevator speeches, value statements,
impact statements, presentations, and other documentation so
they can deliver the right versions of governance messages to
the right people with the right level of detail. They employ
stakeholder participation matrices to make sure stakeholders
arent overlooked and that the right people get their part of the
message in the right sequence. And, they employ email
templates and templates for presentations and reports.
Benchmark
At the industrys first
Data Governance
Conference, in
Orlando, Florida
USA in December
of 2006, leaders
of successful
Data Governance
programs declared
that, in their
experience,
Data Governance
is between
80% and 95%
communications!
They uniformly
expressed surprise
at the amount of
written and verbal
communication
required to
successfully bring
together
stakeholders to
achieve their goals.
Is communication
not your fort?
Thats OK.
Communication
plans and
supporting tools
can be developed
by those with
expertise and then
administered by
your program staff.
Page 19 of 20
Getting Started
What type of Data Governance is right for your organization? How
should you begin your efforts?
Its tempting to start right in designing your Data Governance and
Data Stewardship organizational bodies, assigning roles and
responsibilities, and developing policy. And its true: this is
important work.
But remember that such program design work is actually your fourth
step of seven in implementing a Data Governance program.
Establishing your focus and value proposition should be your first
priority. Be sure you understand how your efforts can contribute to
your stakeholders need to increase revenue and value, manage cost
and complexity, and ensure survival through attention to risk,
compliance, and vulnerabilities.
Not sure? Thats understandable as we said before, its hard to see
the forest when youre one of the trees! Consider asking someone else
from within your organization (or from without) to help you
understand your value statement and to develop a plan to clearly
and unambiguously communicate that value.
When you can clearly describe your organizations data-related
problems, how youre going to address them, and how success can be
measured, then youll be ready to reap the benefits of
a value-based Data Governance program.
Page 20 of 20