Copyright 2009, Information Builders. Slide 1 Why Data Quality Matters Chris Bevilacqua iWay...
-
Upload
franklin-fisher -
Category
Documents
-
view
216 -
download
1
Transcript of Copyright 2009, Information Builders. Slide 1 Why Data Quality Matters Chris Bevilacqua iWay...
Copyright 2009, Information Builders. Slide 1
Why Data Quality Matters
Chris BevilacquaiWay Solutions Architect
Copyright 2009, Information Builders. Slide 2Copyright 2007, Information Builders. Slide 2
Data Quality is…
The
Cornerstone of
Accurate BI
Copyright 2009, Information Builders. Slide 3
Stated another way:
Copyright 2007, Information Builders. Slide 3
BI on bad data is a disaster!
Copyright 2009, Information Builders. Slide 4
What is Data Quality?
Data quality measures:
• Accuracy
• Completeness
• Consistency
• Uniqueness
• Timeliness
• Validity
Copyright 2009, Information Builders. Slide 5
The Real Cost of Bad Data
More than 25% of critical data within large businesses is somehow inaccurate or incomplete. 1
Poor data quality costs the typical company at least 10% of revenue; 20% is probably a better estimate. 2
Poor quality customer data costs U.S. businesses $611 billion a year in postage, printing, and staff overhead.” 3
Your Company’s Revenue = $DDD/yrWhat is the cost of your bad data?
Copyright 2009, Information Builders. Slide 6
Here are a Few Approaches to Solving the Data Quality Problem….
Copyright 2009, Information Builders. Slide 7
We’ll get there…someday…
Whether or not you believe in climate change, the business will change…
Copyright 2009, Information Builders. Slide 8
But when (really) will ALL your data be in the “One System”?
We have One System…..To Rule Them All!
Copyright 2009, Information Builders. Slide 9
It works, but ends up being pretty messy…and did we mention change?
We’ll build it ourselves…what we need, when we need it.
Copyright 2009, Information Builders. Slide 13
iWay Data Quality Center
1. Best practices are “baked in”
2. Built from the ground up for data quality
3. True real-time capabilities
4. Robust international support
5. Pluggable into any topology
ProfilingAnalysisParsingStandardization
ValidationPattern MatchingEnrichmentRecord Matching
LookupsScoringMerging/UnificationDe-duplication
Copyright 2009, Information Builders. Slide 14
Address Identifier Alter Format Apply Replacements Apply TemplateCharacter Groups Analyzer Column Assigner Condition Convert Phone Numbers
Create Matching Value Create Postal Address CZ Data Format Changer Data Quality Indicator
Dbf File Reader Dictionary Lookup Generator
Dictionary Lookup Identifier Dictionary Lookup Reader
Erase Spaces In Names Excel File Reader Excel File Writer Extract FilterFilter Fixed Width File Reader Frequency Analysis Generate Fake RC CZGet Birth Date From RC CZ Get Person Type CZ Group Aggregator Guess Name Surname
Incremental Manual Override Builder
Indexed Table Reader Integration Input Integration Output
Intelligent Swap Name Surname
Jdbc Reader Jdbc Writer Join
Kill Unsupported Characters
Lookup Lookup Builder Lookup Reader
Manual Override Builder Matching Lookup Reader Matching Values Multiplicative Guess Name Surname
Multiplicative Lookup Multiplicative Pattern Parser
Multiplicative Regex Matching
Multiplicative Validate Phone Number
Multiplicator Pattern Parser Profiling RC Validator CZRVN Validator Record Counter Regex Matching Relation AnalysisRepository Key Converter Repository Reader Repository Writer RepositoryReader/2
RepositoryReader/3.0 RepositoryReader/3.5 Representative Creator SIN ValidatorSQL Execute SQL Select Scoring Scoring SimpleSelective Matching Lookup Reader
Selective Transliterate Selector Simple Group Classifier
Sort Split Out Trailing Numbers Splitter Statistics
String Lookup String Lookup Reader Strip Titles Swap Name SurnameTable Matching Tail Trashing Text File Reader Text File WriterTokenizer Transform Legal Forms Transliterate TrashUIR ADR Generator CZ Unification Unification Extended UnionUnion Same Update Gender Update Person Type By IC
RC CZValidate Bank Account Number CZ
Validate Birth Number UA Validate DIC Validate Email Validate IC CZ
Validate ID Card Validate In RES Validate Phone Number Validate RZ CZValidate RZ SK Validate VIN Value Replacer Web LookupWord Analyzer Xml Writer
40+ DQ Focused Objects
Copyright 2009, Information Builders. Slide 15
Real World Use Case
The Goal• Services organization supporting the airline industry sells decision support information to
the industry members.
The Challenge• Data Quality was adversely affecting the customer base satisfaction
• Data Quality was impacting new revenue generation opportunities
The Strategy• Profile analysis according to specific business validation rules
• Monitor rolling 13 month window comparison of monthly data profiles
• Accumulate and report analysis to data providers
The Benefits• Improves customer satisfaction and confidence in the information
• Increases reliability of the information as new data sources are added
• Documents and audits quality-control processes for customer review
• Reduces the dependency on human resources to detect and correct data quality issues
Copyright 2009, Information Builders. Slide 16
Real World Use Case
The Goal• Manufacturer Pharmacy Automation and Nursing Automation Platforms
• Data Synchronization and Data Quality
• Address parts and technicians being sent to wrong facilities
The Challenge• Four different systems
• Two MS SQL Server, 1 Oracle, 1 Progress databases
• Product, Shipping, and Customer data out of sync
The Strategy• Standardize, Cleanse Data across all systems
• Match and Merge Data and maintain ongoing integrity
The Benefits• Deliver Dynamic Single Views
• Prepare for an ultimate MDM initiative
Copyright 2009, Information Builders. Slide 17
Impact of Data Quality Address Data
3658; 36%
2727; 27%
3799; 37%
Verified OK
To be checked manually
To be corrected manually
36 %Naturally Correct
64 %Manual Attention
Copyright 2009, Information Builders. Slide 18
3 %Manual Attention
3658; 36%
2727; 27%
3492; 34%
307; 3%
Verified OK
Standardized & Verified OK
Corrected automatically
To be checked manually
Impact on Data Quality Address Data
61 %Automated Cleansing
36 %Naturally Correct
+
Copyright 2009, Information Builders. Slide 19
Real World Use Case
Goal • Performance Management
• Business Intelligence
• Change Management Process
The Challenge• 100 Locations
• 14 Systems with out-of-sync master data
The Strategy• Cleanse, Standardize, Match
• Master Data Management – Directorate, Borough, Site, Service Type, Service Point, Team, Staff, Patient
• Master Data Governance Workflow
The Benefits• Dynamic organizational change to support strategic initiatives
• Complete visibility into performance of organization vs goals
Copyright 2009, Information Builders. Slide 20
Real World Use Case
The Goal• Major hospital group is building a Master Patient Index
• Need to bring in acquisitioned systems
• Cleanse, Standard, Deduplicate
The Challenge • Previously manually processed by hiring temporary staff
• Current phase projected to take temporary staff of 20 over 18 months
The Strategy• Automate the cleansing, matching and merging business rules
• Data Stewardship provides human oversight to automated process
The Benefits• Identifies the duplicate records according to very complex business rules
• Reusable rules for future phases
• Significantly reduced project time – from 18 down to 4 months.
• Over 400% ROI projected
Copyright 2009, Information Builders. Slide 22
Your Data
Data Quality Challenge
Data Quality Profile
Data Quality Profile
No COST,
No kidding!