There are many approaches to Data Quality Management today. This is an introduction to an approach we have developed predominantly within Financial Services which we believe yields fast, cost effective results for our clients.
Our approach is inspired by Lean Manufacturing and uses the latest Data Profiling tools and techniques to achieve faster results than traditional approaches to Data Quality Management.
2. Data to Value
Nigel Higgs
Experienced Information
Management Consultant
who has led practices within
Financial Services,
Insurance & Government.
James Phare
Former Head of Data
Architecture & Information
Management at Man Group
plc.
Data to Value is a newly formed
consultancy specialising in all aspects
of Information Management
Specialists at bridging technology &
business understanding gap within
Financial Services
We offer a fresh approach compared to
other large consulting firms
Our principles are founded on lean &
agile approaches with our clients’
requirements & priorities at the
forefront of our work
3. Traditional approaches to DQ
Management
Do nothing / ignore (not
recommended!)
Enterprise data quality
programmes
Ad hoc using traditional tools
(Excel etc)
Outsource responsibility to third
party
4. Lean Approach to DQ Management
“Focused on minimising the
allocation of resources to any
activity not linked to creating value
for the customer”
Key features:
Early adopters, prototyping & Minimum
Viable Products (MVPs).
Wide IM skillsets & cross functional
teams.
Smaller batch sizes / more frequent
iterations.
Hypotheses, metrics & ‘validated
learnings’.
Build-Measure-Learn cycle & Pivots.
5. The Process
Data loaded & profiled, discovery
process begins
DQ Rules used to test hypotheses,
capture metrics & quantify defects
Results presented for review
Defects prioritised, managed & resolved
Results inform both recurring & ad hoc
responses to DQ issues by Business
data owners
Discover
MeasureManage
6. Scorecards with dimensions Key Performance Indicators
(KPIs)
80%
85%
90%
95%
100%
Integrity
Completeness
UniquenessConsistency
Conformity
75%
80%
85%
90%
95%
100%
May June July August
Overall Quality Completeness Consistency
Conformity Uniqueness Integrity
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
100,000
May June July August Dept A Dept B Dept C Dept D
Data Quality Scorecard – Product Data May 2013
Commentary:
- Data coverage of priority field
A increased from 30% to 80%.
- Tier 1 incidents down 30%
- Data Dependency X now live.
- Data Maintenance
requirement reduced by 30%
over 3 months to date.
- 2 additional power users
trained in DQ tool.
- 2 additional Data Stewards
within Product Master Data Set.
Status =
Green
May DQ dimensions
Latest dependenciesData Maintenance activity
Monthly DQ dimensions
Aimed at
- Sponsors
- Decision makers
Aimed at
- Data practitioners
- Highly dependent stakeholders
Tailored presentation of results
8. Lean Information Management Specialists.
Data to Value Ltd
42-44 Bishopsgate
London EC2N 4AH
T +44 (0)208 278 7351
www.datatovalue.co.uk
Nigel Higgs
nigel.higgs@datatovalue.co.uk
James Phare
james.phare@datatovalue.co.uk