MDM Best Practices
MDM Best Practices
MDM Best Practices
MDM
DM
A Supplement to
www.DataDelta.com
“At DataDelta we don’t make the match engine,
we make your current match engine work better!”
Master Data Management
9 MDM
Milestones
Sponsors
02 DataDelta
09 Exeros
11 D&B
12 DataLever
MDM
Changing Landscape
12
Successfully delivering projects on time has always been a challenge. Project
management methodologies are well documented and continue to evolve to address
the challenges. The Project Management Institute (PMI) and body of knowledge guide
(PMBOK) are excellent sources of reference material on the subject. With this in mind,
six key points are vital to the success of your MDM project.
5
and performs tasks such as implementing processes and training users
of the system. The change agent leads the change project and requires a
mixture of good project management skills to keep projects on track and good
leadership skills to motivate and evangelize for the project with passion.
6
The main activities break into the following steps: Mapping and analyzing
the “as is” processes, designing the “to be” processes and implementing
the reengineered processes. Following implementation, BPR enters a
cycle of continuous improvement through ongoing measurement and
review. At the core of BPR are processes.
3
scoping step with principles from change management and BPR.
The crucial steps in the lifecycle of an IT implementation project are
typically: scope, planning, execution, control and closure.5 Flexibility,
common sense and communication within this framework are critical for
successful project delivery.
MDM will shift data ownership to the During the scoping phase, the systems and their requirements or
business side of the organization, putting features are identified and the business requirements are gathered. The
focus on a data methodology. project deliverables are broken down into manageable tasks, resource
Prior to MDM, data needs were typically assignments and milestones. These projects are typically
fulfilled by each department maintaining led and managed by IT.
their own silo of information and/or
Both
requesting a snapshot of data from a data
warehouse maintained by IT. Centralized
data was owned and maintained by IT. a business sponsor and an ROI case for the project
Data needs often lack a methodology for are needed for continued executive support during the lifecycle. A steering
preparation and analysis. But, at the core committee working at a strategic level guides the project, and project
of the methodology is data. management focuses on the execution of the plan within the constraints of
time, resources and scope. At the core of IT implementation is execution –
getting the project completed on time. Implementing MDM is still relatively
new to organizations, compared to implementing ERP or CRM.
continued
a
IT Application
Implementation
a
Business
Process
Fig Reengineering
ure
1:
a
Data
Needs
Fra MD
me M Change
wo Management
rk
Data Governance Council – PMO – Data – Processes – Standards
“ Great Networking!
Overall, a great
“ A fantastic conference experience!” “ This conference is a
on what is becoming James Cisler, great way to see what
the “next big thing” Manager, Master Data, Stryker vendors are doing.”
in IT 0
”
Robert Zegrac,
Mark Evans,
Senior CDI Consultant,
Group Manager, National City Experian
Presented by:
Excerpted from “Milestones on the MDM Roadmap for 2008-09” by the MDM Institute.
continued from page 8
Overcoming Complex Data Challenges offers a graphical interface and “out of the box best
The difficulty in showing “quick success” is that practice” project accelerators dramatically increases
the data and necessary business rules/logic are staff productivity and ROI. Offering an interface that
very complex. Typically, the integration of data supports the efforts of both business users (who
and ensuring its’ fitness for business use consume have unmatched domain expertise) and technical
50-70% of project time and budget. Identifying and users (who understand technology capabilities and
resolving entity information relating to customers limitations) accelerates efforts toward information
and products is based upon extensive reference sharing and collaboration.
data and knowledge bases that are not easily Offering these sophisticated yet easily used
replicated. The ability to understand “hidden” capabilities in a very high performance platform
information and relationships buried in data is a ensures that current or future data volumes will not
daunting computational challenge. overwhelm your MDM system. Technology that can
In addition to correctly identifying information, be deployed in batch mode for initial master data
“fuzzy matching” of information within or across “build” and then further deployed as a common
sources is a very complex and difficult task. Matching data service for universal access, extends value.
different sources- or even matching within a source- Accommodating customer information whether
is made difficult by inexact representations of the B2B or B2C, as well as product information is
same thing or similar representations of different very useful. Properly handling address information
things. Normal database joins can’t overcome these for correct billing or marketing efforts, as well as
natural language variations. Having sophisticated geospatial information for tasks like tax or risk
algorithms that overcome these problems, along assignment, or “find closest” applications, adds
with other common errors like nicknames, alternate further value to corporate data.
firm names, different spellings, and data entry errors
is a key attribute of a reliable and extensible MDM About DataLever
strategy. With an 80%+ win rate in competitive bakeoffs,
Performance and accuracy of matching algorithms and performance up to 25X that of competitive
is paramount- missed matches or false matches technology, DataLever is a leading provider of
create risk, miss potential value, and significantly business value focused MDM technology.
reduce the trust of business data owners. Matching Companies with large data volumes and/or
is one of the hardest yet most critical attributes complex business rules like AAA, AIG, Acxiom,
in creating “trusted” master data. Thus matching Equifax, Experian, the US Department of Veterans
accuracy and capabilities should be a critical Affairs, and Time Warner Cable rely on DataLever.
evaluation and selection criteria. We provide a free data test/POC to show how
quickly the most accurate view of your data can
Functional Capabilities drive your success. •
In addition to entity identification and fuzzy
matching, there are a number of other key functions
that create greater business value with master www.datalever.com
[email protected]
data. Providing a single integrated platform that 703.956.6436