Open Framework, Information Management Strategy & Collaborative Governance | Data & Social Methodology - MIKE2.0 Methodology
Wiki Home
Collapse Expand Close

Members
Collapse Expand Close

To join, please contact us.

Improve MIKE 2.0
Collapse Expand Close
Need somewhere to start? How about the most wanted pages; or the pages we know need more work; or even the stub that somebody else has started, but hasn't been able to finish. Or create a ticket for any issues you have found.

Data Quality Management Business Case

From MIKE2.0 Methodology

Share/Save/Bookmark
Jump to: navigation, search

Data Quality Management business case is performed by organizations planning to embark on a Data Quality Improvement and Reengineering initiative. This summarizes the cost/benefits of conducting a data quality program. The below business case was done for a banking organization and can be used as a reference for creating business cases for other organizations/projects.

Contents

Executive Summary

Project Goals

Assess the Magnitude of the Current Data Quality Issues

  • Utilize tool-based assessment of source data to quantify the magnitude of data quality issues
  • Conduct survey to prioritize improvement initiatives

Establish Enterprise Data Quality Management Strategy & Roadmap

  • The Strategy prioritizes Data Quality Management capabilities and makes recommendations regarding what actions should be taken
  • The Roadmap documents how an organization should go about implementing the recommended Strategy

Establish Business Case

  • The Business Case demonstrates the value that can be expected by funding the Data Quality Management Strategy as outlined in the Roadmap document

DQM Drivers

Key Quality Issues Observed

  • Data Inconsistencies Exist Across Systems
  • Staff Time Wasted Massaging Data
  • Fragmented View of Customer Exists
  • Customer Records Are Duplicated
  • Accuracy Issues Exist for Key Data Elements

Impetus to change

  • Drive Down Costs
  • Improve Competitive Position
  • Meet Regulatory Requirements
  • Enable Business Process Flexibility

Desired State

  • Key Data Elements in Synch Across Systems
  • MIS Staff Spends Time Analyzing not Verifying
  • Integrated View of Customer Across the Bank
  • Customer Records are Unique
  • Required Accuracy Levels are Consistently Achieved

Business Case Background

The business case is organized around the three phases which correspond to the DQM Strategy

  • Phase 1: Organize and Define the Data Quality Program
  • Phase 2: System-Enable the Data Standards
  • Phase 3: Optimize Data Integration

The focus of the DQM project has been on the credit subject area of the warehouse, however many of the benefits articulated in this business case will have a positive impact on the finance subject area and enterprise-wide.

The benefits are derived from diverse sources such as improved risk management, regulatory compliance, increased sales, and improved productivity There are many additional benefits derived from improving data quality which have not been quantified.

Business Case Summary

Recommend committment to Phase 1 now.

DQI-Supporting Asset-Business Case 1.jpg

Provides the flexibility to realize 5 Year NPV = $3.682M ROI = 62.7% Breakeven month for entire commitment if and when done (in 3 separate phase commitments) = 35

DQM Rollout Decision Option Tree

Using the cost, benefit and risk analysis, this decision tree supports our recommendation to commit to Phase 1 now. We can use the risk factors and benefits realization to reassess the Phase 2 commitment when the time comes.

DQI-Supporting Asset-Business Case 2.jpg

Annualized Cash Flows and NPV

DQI-Supporting Asset-Business Case 3.jpg

Business Case Assumptions

The costs have been aggressively estimated and the benefit estimates are conservative making this a pessimistic scenario The part-time participation of Business Resources as Sponsors and Stewards has been factored into the project costs. Because of the nature of Data Governance, it is assumed that existing staff will fill these roles with no additions to staff. The $3M+ FTE costs in the business case reflect no net new adds to staff.

Resource Valuation

  • Business Resources have been valued at $33/hour or $55/hour depending on the level of the resource
  • IT Resources have been valued at $91/hour
  • Consultants have been valued at $300/hour to approximate the rate plus expenses

Benefit Detail

Benefits by Phase

DQI-Supporting Asset-Business Case 4.jpg

Improved Development Productivity

At least 5 FTE are dedicated to major data-intensive projects. A 10% productivity enhancement will be realized through improved metadata at each phase of the Data Quality Program. This productivity improvement would positively impact all future enterprise projects which rely on EDW data. It is assumed that the Bank will do two projects per year after Phase 3. The value of this benefit is $597,000.

Improved Relationship Reporting

Common Customer Identifiers (CCIs) can be changed, severing historical reporting continuity. The impact of this issue is significant because certain organizations relies on the CCI as a primary key for linking customers with their accounts and other customers. Credit modeling and analysis depends on continuity over periods of five to seven years. The calculation of customer profitability will be favorably impacted by the correction of this data quality problem. The organizations collectively would save 0.15 FTE if they did not have to research and repair customer relationships. A productivity savings of $28K would be generated.

This quantified benefit does not begin to reflect the actual amount of effort expended in many areas across the Bank researching fragmented customer relationships.

CRA Reporting Accuracy, Efficiency & Cost Reduction

System edits around Annual Sales and Customer-to-Account relationships would improve the accuracy of CRA reporting. The OCC will accept a 5% error rate at the field level but no more that 10% errors overall. Should the Bank fail to obtain a Satisfactory CRA rating, the Bank would be precluded from acquiring other financial institutions and would not be eligible to bank government entities. This benefit calculation assumes a 5% risk of not obtaining a Satisfactory CRA rating. Improved data quality reduces that risk by 1%. In order to address the increase in work load, CRA Reporting engages contractors to assist during peak reporting season (at year-end).

DQI-Supporting Asset-Business Case 5.jpg

$40K is spent annually on consultants. The current full time staff could handle the work load with less outside assistance if they were not as engaged in data quality verification, reducing consulting expenses by $25 - $30k annually.  Total Benefit is $360K

Consolidated View/Edit of Key Data Elements

Currently Corporate and Middle Market RMs (80 FTE) spend time each month on some data quality review of their portfolios and on remediation of data errors. The current rate of customer acquisition in commercial banking is 15 Corporate and 30 Middle Market customers per quarter. The profit from established middle market customers averages $25k/month. If the data review and remediation process was streamlined with consolidated view and edit capabilities provided through system interfaces, RMs could spend more time selling. At a rate of 1 additional new customer every other month, the Bank would realize a significant increase in customer profitability over time as a result of improved data interfaces. This benefit assumes that new customer profitability gradually reaches $25k/month over 12 months. Over five years, this benefit is worth $10M.

Credit Examiner Efficiency

The Credit Examiners (7 FTE) have identified poor understanding of certain credit policies and the lack of departmental QA processes as key factors contributing to poor data quality. They estimate that 10-15% of their time is spent on data quality related research and discovery activities. A 5% FTE savings for the team would be realized as a result of credit policy refresher training and a 5% FTE savings would be realized as a result of departmental QA process implementation. Total Benefit is $303K.

Returned Mail Reduced/Eliminated and Consumer Marketing Effectiveness

Approximately .5% of DDA statements are returned because of undeliverable addresses. A similar return rate is experienced by consumer marketing using the same addresses. This returned mail is handled by 1.5 FTE in BOS. CFSG marketing mailings go to a smaller audience but have a higher return rate (3-4%). ESB Administration handles this returned mail. One FTE savings in BOS could be be realized by periodic cleansing of addresses.

In addition to the handling costs of returned mail, the missed sales opportunity on consumer marketing mailing can be calculated. The average response rate to solicitations is 2%. In conjunction with the returned mail rate of .5%, we can calculate the missed opportunity cost of returned marketing mail. This is based on an assumption of $100 annual profit per product sold. Benefits total $257K.

Improved Data Administration Productivity

Automated cleansing of Customer data in CIF would increase the number of automatic merges which can be performed because names and addresses would be standardized. The Data Integrity team (6 FTE) currently has a 2 week backlog of officer requests and cannot re-assign the portfolios of officers who have left the bank. Approximately 9,000 commercial customers and 7,000 commercial prospects who used to be managed are now unmanaged. A .5 FTE savings would be realized by reducing the number of duplicate customer records that have to be manually merged. This resource could attend to the un-assigned portfolios and improve the bank’s cross sell and prospect conversion rates. Total benefit is $143K. In addition, the team could leverage the same cleansing and matching routines to reduce duplicate prospect records yielding additional productivity.

The enterprise benefit of data cleansing and customer linking has not been quantified.

Duplicate Record Reduction

Duplicate customer records could be reduced as a result of restricting new profile creation to the Data Integrity team. The FTE is engaged to identify and merge suspected duplicates. To eliminate duplicates a manual merging process has to be performed. Together, these areas would realize a 0.15% FTE increase in productivity as a result of this change. Total benefit is $40K. The reduction of duplicate records in benefit ICRS.

Basel Compliance Capital Benefit

A decision has been made to attempt to qualify for the AIRB (Advanced Internal Ratings-Based) approach to calculating capital requirements when Basel II requirements go into effect in 2006. Qualification for the AIRB is based on demonstrating to the regulators that the Bank has all the data necessary to support the capital requirements calculation and certifying the accuracy of that data. If the Bank qualifies, in 2006 the capital requirement will be 90% of the current amount and 80% in 2007. Beginning in 2008, if the Bank can meet the regulatory requirements, the Bank will be able to calculate its own capital with no minimum requirements. Using this approach, the Bank can free up approximately $442 million in 2006 and $884 million in 2007. The after-tax cost of capital is 9%. The decision tree shown on the next slide assumes that the initiatives currently underway in Credit Portfolio Management to accumulate the necessary historical data in order to qualify for the AIRB approach will continue. The benefit of embarking on the Data Quality Management program is derived from a 10% improvement in the likelihood of the Bank qualifying for the AIRB approach. The benefit of qualifying for the AIRB Approach begins in 2006 and continues to grow each year through 2008. Total benefit claimed is $5M.

Capital Charge Benefit from DQM

The value of the DQM Project in qualifying for AIRB is shown on the top of the decision tree. The other Basel-related qualification activities are depicted on the bottom half of the tree. The DQM Value is the difference between the top ½ of the decision tree (doing DQM along with other initiatives) and the bottom ½ of the decision tree (just doing the other initiatives)

DQI-Supporting Asset-Business Case 6.jpg

Correcting Invalid SIC Codes

3% of the commercial loan portfolio has invalid (9999, null) SIC Codes. The industry multiplier that is applied to allocate capital an average of 1.65% higher for loans without a valid SIC Code. As a result of a Phase 1 Quick Win clean-up, a benefit of approximately $483K will be realized. After Phase 2 system edits are implemented, this benefit will continue. To calculate this benefit, it was assumed that the affected loans have smaller balances and are scrutinized less often.

ICRS Timeliness

Embarking on the DQM program immediately can positively impact the timely completion of the first phase of ICRS by mitigating delays caused by poor data quality. This benefit assumes that 10 FTEs will be involved in testing and reconciliation remediation. For each month that ICRS is delayed, the cost of those resources will be expended. Preventing a one month delay would yield a $137K benefit. The DQM program will also improve the value and adoption of ICRS as a result of the higher quality of data which will be presented to end-users. The DQM program positively impacts the timely delivery of the finance subject area of the EDW2.

Intangible Benefits

The following DQM program benefits have been identified but not quantified

Reputation: Positive Impact on the Regulators’ impression of the Bank’s commitment to improving data quality Downgrade Triggers: When a loan downgrade is performed, ideally a review of the customer’s deposit accounts, other loans and non-loan credit products should be performed. De-duplication facilitates this review because all of a customer’s accounts are linked to the same customer record. Credit Exposure: A more accurate view of customer and counter-party exposure will be achieved through de-duplication; comprehensive participation exposure will be achieved through obligor linking Reduced Turnover Exposure: Currently the Bank relies on Subject Matter Experts in the MIS domain for new projects and day-to-day operations. The development of level 1 metadata reduces the risks associated with the departure of key experienced staff. Householding: The current householding algorithm (generated in MCIF) is not as sophisticated as Marketing and Finance would like, resulting in an overstatement of the number of households. An improved customer linking algorithm would aggregate customers more accurately and allow for more targeted marketing. EDW2 Adoption: The perception that data integrity in the warehouse is high will improve the likelihood that data marts will source information from EDW2, reducing or eliminating dependence on source-system extracts.

Cost Detail

Phase 1 costs

DQI-Supporting Asset-Business Case 7.jpg

Phase 2 costs

DQI-Supporting Asset-Business Case 8.jpg

Phase 3 costs

DQI-Supporting Asset-Business Case 9.jpg

Ongoing costs

Staffing DQI-Supporting Asset-Business Case 10.jpg

Software Maintenance Estimate: $300K Hardware Maintenance Estimate: $120K

Wiki Contributors
Collapse Expand Close