The Importance of Data Quality Management and Data Cleansing for Banks

Updated On : February 2016  |  by Amol S. Khanvilkar

Dramatic transformations in banking have been driven both by new regulatory requirements and the technological advancements that aid banks in meeting and exceeding such requirements. One of the largest challenges faced by banks today is managing data – both for regulatory requirements as well as to gain meaningful insights.

Businesses driven by modern technology operate through a large number of channels which in turn generate large volumes of data that modern banks must support. New technologies such as ERP, SCM, and CRM systems have been introduced to help support the needs of such organizations. These generate almost unheard of volumes of data that modern banks must manage and ensure the quality of.

The data that banks are concerned with typically contains large numbers of business transactions and records and needs to be accessed for a various number of banking functions instantly from throughout their banking networks. Couple this with the strict nature of regulations banks must adhere to with their data in comparison to other business sectors, and the true size of the need for Data Quality Management and Cleansing for Banks becomes evident.

There are quite a few significant challenges banks must be aware of, focus on, and overcome to ensure proper Data Quality Management. Perhaps the most key are:

• The sheer volume of data
• Securing data at all times
• Maintaining all statutory and regulatory requirements
• Interfacing with legacy applications securely and efficiently

So what strategies and tools are available to banks to overcome these challenges as they work to ensure regulatory compliance and proper Data Quality Management? To begin with, a firm understanding of what Data Quality is and what managing it in the Banking sector looks like.

Data Quality Management for Banks

Certainly every business and IT department must be concerned with the quality of the data it maintains. However, the traditional needs of quality management are exacerbated by the unique circumstances of the banking sector described above (volume of data, regulatory requirements, legacy systems, etc.).

Here are few more significant reasons that Data Quality is of utmost importance for banks:

• The every evolving need for Risk Management Applications creates an even more complex network of data and puts an even greater need for the accuracy of data.
• The explosive growth of ecommerce has led to the creation of several new sources of revenue for account holders.
• The regulatory guidelines banks must face all over the world are continually evolve and become even stricter. What works today may not be adequate in a few short years forcing banks to be on the forefront of data quality and security efforts.

Defining Data Quality

The quality of data, in this context, can be understood as it’s suitability for meeting the needs and requirements banking institutions require of it. To be clear, data doesn’t have to be perfect, but rather, it needs to meet the requirements of whatever system utilizes it or those systems return inaccurate results. To help ascertain whether or not data is high quality, a number of specific factors are taken into account:

• Data Integrity
• Data Completeness
• The Data’s Accessibility
• The Data’s Timeliness
• Accuracy of the Data
• Validity of the Data
• Integrity of the Data

There are a number of causes that lead to a loss of data quality which will be covered more thoroughly in the next section. These include duplicate records, missing data, incorrect data, and even errors created during data entry.

Cleaning Data

So how do banks manage the quality of their data over and above simply trying to keep a better watch over these issues? There are a couple of employable strategies:

• Locate and correct inaccurate and defective elements and values like misspellings and mistyped number values.
• Standardize data by modifying it to uniformly confirm to standards that make using and understanding it easier and more effective. This can be accomplished by matching and merging records within a file.
• Use filtering technique to catch duplicate, nonsensical and even missing data

The best place to clean data is always the source system or application. If this is not available, other options are:

• During an ETL
• In a Data Warehouse
• In a Staging Area

Where Unclean Data Comes From

Misleading, missing, duplicate, or otherwise unclean data can come from quite a number of sources. These include but are not limited to:

Interfacing and integrating with other systems and databases across the globe. Systems are set up differently in different parts of the world, miscommunication happens between systems just as it does between speakers of different languages.

Any paper documents anywhere in the data chain can easily be the source of error as they require manual input into electronic systems.

Any changes to the account holder’s information that needs to be shared across different applications and systems within the banking network. For example, if an account holder gets married but the name change is not carried over to all accounts automatically.

Often information from different places such as call centers is incomplete as operators often have to enter it in a hurry which requires them to condense or leave out details.

Any data from third-party partners or systems that has errors in it could enter automatically and be incorrect. There are constant mergers and acquisitions in the banking industry. This constantly requires reintegration of data which can lead to duplicate entries, missing entries, and even corrupted data.

The Benefits of Cleaning

Cleaning Data and managing it to maintain quality provides banking institutes a number of advantages. Not only does it increase confidence in reports generated from the data, it ensures decision making is supported by accurate information.

Additionally, having systems in place to account for duplicate and unclean data automatically dramatically reduces the amount of time accounting staff must dedicate to such tasks. Additionally the amount of communication generated and transmitted internally and externally through banking networks about such incorrect data would be eliminated as well.

Clean data means effective business and increased profitability for the bank and its account holders by eliminating common mistakes like duplicate and missing mailings that can be directly related to unclean data.

To ensure the best implementation, data cleaning solutions should be done proactively and never after a failed or bad campaign. Here are three quick steps to help ensure a successful implementation of a data quality solution:

Step 1: Hire an external IT consultant to conduct a database audit to ascertain current data quality.

Step 2: A Data Quality Solution should be implemented before or simultaneously with any other planned data management solutions like data warehousing.

Step 3: Assemble a team of analysts, IT personnel, and experts from different domains or application areas that rely upon clean data should be formed to oversee the data quality management.

A partnership with a trained and experience technology consultant firmed is highly recommended for any bank or financial institute looking to proactive implement a data quality solution.

Nelito has a tries and tested Data Quality solution which helps banks in UCIC (Unique customer Identification) programs as well as in data enrichment and various other related programs.