data governance and data quality stewardshipdamaiowa.org › wp-content › uploads › 2016 › 06...
TRANSCRIPT
Data Governance and Data Quality
1
Stewardship
Agenda
Discuss Data Quality and Data Governance
Considerations for future technical decisions
2
3
Portal Embedded InfoApps™
Applications Legacy Systems Relational/Cubes Big Data Columnar/In Memory Unstructured Social Media Web Services Trading Partners
Integration
SocialHot
BadFeedback
Predictive Analytics
Sentiment and Word Analytics
Search Location Analytics
Mobile Write-Back
Data Discovery Reporting Dashboards Casting and Archiving
Active Technologies
High-Performance Data Store
Data Quality
Data Governance
Master Data Management
Batch ETL Real-Time ESB
Integrity
Intelligence
4
5
Tell yourself they are the same, it doesn’t matter!
Which ones are bigger?
Data Value Chain
Connect
Data
Application
E-Business
Legacy
SaaS
Big Data
Data
Move
Batch
Transactional
Event Driven
SOA
Automate
Orchestrate
Fix
Profile
Clean
Enrich
“DQ Firewall”
Relate
Master Data
Organize
Synchronize
360 View”
Govern
Monitor
Visualize
Alert
Remediate
Report
History
Business Intelligence
Dashboards
Analytics
Ad Hoc Reports
Enterprise Search
Mobile
Visualize
Predictive
Social Intelligence
Performance Mgt
Business Value
Integration Integrity Intelligence
Anyone Facing these Challenges?
7
Difficult to produce accurate customer
count
Not clear who
“owns” data
Different answers for the same
question
LOB “manage” their own
data in Excel
Duplicate data across
systems
Resources tied up
researching and fixing
data issues
Errors in processing data due
to incomplete
data No single version of the truth
Why Information Management
Barriers to Information Management Address disparate, dirty and timeliness of data
Data Spread Across Too Many Apps and Systems 67%
Multiple Versions of the Truth 64%
Data Not Timely Enough 60%
Data Not Clean Enough To Use 58%
Technology Not Able to Meet Needs 57%
Source: Ventana Research Information Management Benchmark Research
Data needs to be treated as a Business Asset
9
Gartner has researching the concept that information is an under-managed, under-utilized asset because it's not a balance sheet asset.
Gartner – 25% of critical data is flawed
Not ALL data should be managed equally
How can we think about Data Quality?
Potential Energy = m * h * g
10
There will not be a test after the talk!
How can we think about Data Quality?
11
Dat
a Q
ual
ity
Time
Pote
nti
al o
f D
ata
Optimal Data usage gives continual energy to your organization
Available
Accessible
Authorized
Timely
Efficient
Usable
Defined
Recognized
Structured
Reliable
Consistent
Accurate
Complete
Auditable
What are common aspects of Objective Data Quality?
Data Stewardship enables Data Quality
12
What are common aspects of Subjective Data Quality?
Data Stewardship enables Data Quality
13
Trust
Understandability
Interpretability
Objectivity
Timeliness
Relevance
Benefits of Data Quality
14
Improved productivity with efficient processes Reduced errors HR
Reduced staff for data cleansing tasks Improved productivity of standards based application development IT
Single View of Customer (increases customer satisfaction) Enable better interaction with customers across touch points Ability to cross sell and up sell products and services Accurate Install Base information
Sales, Marketing and Customer Service
Enhanced and Accurate Reporting Efficient Planning and Budgeting with increased granularity Enhanced ability for regulatory compliance Improved decision making based on accurate data
Finance and Corporate
Business Function Benefit of Data Quality, Governance, and MDM
Remediation and Data Stewardship
16
Data stewardship is the management and oversight of an organization's data assets to help provide business users with high-quality data that is easily accessible in a consistent manner.
How is it maintained?
How is it used? How is it
consumed?
People
Process
Technology
Success in Data Quality relies on the harmony of…
People
Process
Technology
Data Governance is supported via Remediation
The “Harmony” is supported via Data Governance
People
Process
Technology
People…
Element of Success: Data Governance
20
Start with a realistic goal
Provide a plan
Be clear with roles & responsibilities
Marketing! Marketing!
Early
Executed
Replicated
Governed
Where is your organization?
Information Management Maturity Quadrant
Wide-spread
Siloed
Sporadic
Systemic
5 Roles of a Data Steward
22
Promote and drive practical governance guidelines throughout an organizations data ecosystem Lead
Work with the business to understand data needs and find better or more efficient processes. Then, the steward can craft appropriate processes to the data uses. Map Per Forrester, “A data steward should drive the implementation and enforcement of requirements for services including, but not limited to, data management, data brokering, customer data integration and data appends. To further drive organizational consistency, the data steward should participate in the vendor evaluation process as necessary to ensure compliance.”
Define
A data steward must keep up-to-date on changes to data-related legislation for external business implications as well as internal communications and compliance.
Be an Expert
Along with daily roles, the data steward should be a point person on the organization’s data evolution. Meaning as new data requirements arise, they must advocate that data governance best practices continue to be utilized.
Advocate
People
Process
Technology
Success in Data Quality relies on the harmony of…
Which Factors to consider
Deciding on the right “Process”
24
Current data skills Company culture
Data reputation Current opinion on data ownership
Maturity of KPI Culture
Reusability of data.
Data Stewardship can be broken into 5 different patterns of allocating the remediation process responsibilities
Allocating Resources to the Stewardship Process
25
By Subject Area
By Function
By Business Process
By System
By Project
• Each Data Steward is in charge of their own subject area. One is in charge of customer and another is in charge of product.
Stewardship by Subject Area
26
Product
Location
Customer
Vendor
Data Management &
Data Governance Processes
Stewardship by Subject Area
27
Cons Focus may be at expense
of broader business benefits (Customer Retention for example).
Size differences of Domains.
Might be difficult to tie Data Steward back to business initiatives.
Pros Boundaries are clear Subject Knowledge Grows
over time
• Each Data Steward focuses on their line of business or department. Such as Marketing or Finance.
Remediation by Function
28
Business Rules & Standards
ERP CRM Inventory FMS
Finance Sales Customer
Service Logistics Marketing
Remediation by Function
29
Cons
Multiple data stewards in different
departments may be managing
and manipulating the same data.
The nature of this model means
that data stewards are rarely
motivated to collaborate across
functional boundaries
Functional data stewardship won’t
work in companies that have
prioritized enterprise-class “single
view” initiatives or consolidation
programs.
Pros
Bounded by the organization
means easier to establish
definitions and rules.
Will be business-savvy and familiar
with the data’s context
They know the team
• Each Data Steward is assigned to a single business process. For example Sales or Enrollment.
Remediation by Business Process
30 Tip : For very mature data-driven organizations
Start End
Start End
Start End
Start End
Data Management & Data Governance Processes
Sales
Enrollment
Procurement
Reporting
Remediation by Business Process
31
Cons
Data ownership is more difficult to
assign. A broader data governance
program is critical for managing such
situations.
Business constituents can get confused.
Consistency around similar types of
data.
In this model, data stewardship is only
as effective as the company is clear
about its processes.
Pros
Extension of exiting processes
Success measurement is more
straightforward
The process oriented model is a very
effective way to entrench data
stewardship.
Data Steward is assigned to the system that they manage the data for. Such as SAP ERP or Salesforce.
Remediation by System
32
ERP
CRM
Inventory
FMS
Data Management &
Data Governance Processes
Tip : This may have caused some of the data quality issues in the first place.
Remediation by System
33
Cons
Business people may equate data
ownership with data stewardship,
thus assuming stewardship to be
“an IT issue”
Data stewards can become myopic
as they maintain the integrity of
the data on their systems
A systems orientation doesn’t
ensure data sharing or
reconciliation.
Pros
IT can take a leadership role
Drives from a Bottom Up approach
Assigning multiple data stewards at
once is more realistic: “each core
system will have a data steward”
becomes an established practice.
• Data Steward is assigned to a project that they will manage the data for. Can be assigned through the PMO office. Examples are a Data Warehouse Implementation or ERP Migration.
Remediation by Project
34 Tip : This can be the fastest way to introduce the role to the organization
Data Management &
Data Governance Processes Project
Management Office
Remediation by Project
35
Cons
“Project” implies ‘ending’
Are skills in house?
Pros
Speed! It is part of the Project and
most organizations can add that as part
of a process easily.
Start with Project then Grow
Clear definition of success
People
Process
Technology
Success in Data Quality relies on the harmony of…
Changing the landscape
37
Data Integration, Quality, and Mastering
38
Typical Historical Approach
End State MDM hub
Data warehouse
Partner interface
Quality process
Operational systems
BI/analytics app
Data Integration, Quality, and Mastering
39
Agile Approach to MDM, Data Quality, and Data Integration
End State MDM hub
Data warehouse
Partner interface
Quality process
Operational systems
BI/analytics app
Traditional in Transition to Modern
40
Fewer use cases
More use cases
Modern Traditional
Hadoop
IoT
Streaming
Virtual DW
Data Lake
OLTP
OLAP
Data warehouses
Data marts
Point-to-point Integration
EII
The Evolution of Integration
41
Hand Coded Integration
ETL Messaging Bus
ESB EAI Hadoop-Based Integration
We Have Some Pretty Simple Problems…
According to a May 2015 Gartner Survey…
26% are deploying Hadoop, 11% in 12 months, 7% in 24 months
49% cite trying to find value as their biggest problem
57% cite the Hadoop skills gap as their biggest problem
To summarize…
Companies are investing in Hadoop, but not sure why
Companies are investing in Hadoop, but don’t know how to use it
42
Big Data Under the Control of Master Data
Hadoop
Master Data Repository
Golden Records
Hadoop can be: • Staging area for application data • Source for mastered subjects • Source for transactional subjects
Master data can: • Provide context to Hadoop data • Establish trust in big data • Guide extraction of Hadoop data
Want More Information?
44