• Cognizant 20-20 Insights

A Path to Efficient Data Migration

in Core Banking
A core banking system implementation can be set up for success
by managing data migration well.

Executive Summary Business Situation

New-age digital disruptors are having a profound Core banking systems not only drive banks’
impact on many organizations across industries, operations but also help to scale up new opportu-
redefining customer expectations and reshaping nities and growth. They have become even more
industry boundaries. The banking industry is crucial as business aligns with digital initiatives,
no exception. It is evolving to safeguard its and given regulatory compliance, M&As and the
existing sources of revenue and find new ones. challenges associated with legacy systems — high
While going digital is the new norm and part of maintenance costs, lack of efficiency and scalabil-
the strategy for many business organizations, ity, technical obsolescence, lack of flexibility.
banks have been lagging behind on their legacy
systems. These systems lack flexible architec- Not surprisingly, 60% of banks are undertaking
ture and are incapable of imitating the digital a transformation of their core banking systems.1
experience provided by companies such as A crucial element of this transformation is the
Google or Amazon. movement of data from the legacy system to the
target system. Data migration exercises involve
Banks, therefore, are now focused on core banking records from different sources and in different
transformation. With this come the challenges of formats. A few banks wrongly assume that the
migrating data from the existing system to the data migration process is a specialized technical/
new platform. The key considerations in any data IT exercise. In fact, business/functional users
migration are data extraction, cleansing, trans- are equally important in the process to ensure a
formation and mapping, and archiving the data. smooth transition. Data migration in core banking
Significant time and money is spent performing is all about the seamless movement of entries,
these activities and hence they must be handled balances, P&L/balance sheet data, customer
with the utmost care. An accurate and timely information, contracts, products, KYC details and
migration is critical for ensuring effective core other forms of financial/nonfinancial data from
banking system deployment and decommission- the source to the target system.
ing of legacy systems.
The data migration process bears significant
This white paper presents an approach and lists risk if not carried out effectively. Indeed, poor
best practices for a successful data migration in data quality can hinder the adoption of the new
any core banking system implementation. system.

cognizant 20-20 insights | november 2016

Challenges • Reconciliation of data: Both financial and
The main problems that banks face during the nonfinancial data from the legacy system must
data migration process are: be properly migrated to ensure data sanc-
tity. With complex business rules and large
• Lack of data knowledge: Data migration volumes of data, reconciliation becomes an
projects fail frequently due to a lack of under- onerous task.
standing of the data in legacy systems. Some of
the reasons for this are: • Lack of flexibility: Change requests that im-
pact the migration process (e.g., addition/
>> Incomplete documentation of legacy systems. deletion of fields at a later stage) have to be
>> Relationship between data not defined analyzed well before being implemented. The
accurately. inability to manage such changes increases
cost and complexity.
>> Dearth of resources that understand the
legacy system data. • Business-as-usual operations: A major risk
>> Assumptions about data structure. that any core banking implementation faces
is its effect on “business as usual” operations.
• Quality of source data: Data quality issues
Any impact on the bank’s customers must be
in legacy systems are one of the biggest chal-
minimized. Work-arounds need to be planned
lenges and reasons for project delays and cost
to support the various digital channels and
overruns. Most banks are not aware of such
maintain operational continuity.
quality issues at the time of embarking on a
core banking transformation. Also, they lack
the expertise to take decisive action on dirty
data in the legacy system. Often data quality An Approach to Migration
issues are not identified till the target system Below we propose a proven methodology and
fails. best practices to mitigate the risks and challenges
involved in the migration process.
• Large volumes: With large volumes of data
comes increased complexity. Huge data vol- Following are descriptions of the key data
umes increase the burden of data governance migration activities for banks engaged in core
and affect data quality. banking transformation. Figure 1 (next page)
• Alignment of business standards and provides an overview.
accounting: Changing business rules impact
banks’ accounting activities. It is therefore Scoping & Planning
critical that the data is mapped in accordance To achieve a successful migration, creating a
with business standards. Different accounting detailed plan is essential before undertaking the
standards practiced in the various geogra- movement of data. The plan sets expectations
phies that the bank operates in must be taken about the complexity of the migration, its timing,
into account. and potential issues and concerns. Scoping and
planning consists of the following subprocesses.
• Mapping of data: Mapping of data is often
metadata-driven. Assumptions-led mapping
• Requirement analysis: This exercise usually
causes significant errors and a high rework involves activities such as business require-
rate. ment gathering and functional specifica-
• Duplicate data: Legacy systems often tion documentation. A thorough review and
contain multiple entries for the same understanding of the requirements helps to
customer. Depending on the bank’s require- accurately determine the scope of the project.
ment and the target system specifications, It is important that all the requirements
the data should be handled so as to avoid captured are well-documented, with sign-offs
duplication or redundancy. from stakeholders wherever required. The
exercise should consider the issues associated
• Short conversion period: The limited time
with the data in the legacy system, security of
available for conversion to the new system
the target system and the new functionalities/
during the cutover period poses a signifi-
data structure of the target system.
cant challenge for implementation. Any error
during conversion delays the implementation.

cognizant 20-20 insights 2

Migration Methodology

SCOPING & DETAILED MIGRATION MOCK RUN MIGRATE RECONCILE

PHASES
PLANNING ➤ ANALYSIS ➤CONFIGURATION ➤ & TEXTING ➤ ➤ & GO LIVE

Source &
Requirement Configure Trial Load in
Target
Analysis Migration Tool Data Load Target System
Profiling

Identify Product Configure Mock Verify

Stakeholders/ Listing Staging Run Target
Team Area Application

Strategy Data Development

ACTIVITIES
Planning Mapping of Migration Testing Target
Scripts

Change
Data
Control
Cleansing
Board

Project
Initiation

TRAINING

Very High Impact Activity High Impact Activity

Figure 1

• Identify stakeholders/teams: The next step is >> Phased: Data is moved to the target system
to distinguish key partners and team members. in a phased manner. For new customers,
It is important that the relevant stakeholder, records are created directly in the target
functional and technical teams are included at system.
an early stage to eliminate the risk of gaps in
the target system. Also, there are numerous
>> Parallel run: Transactions are posted on
both the source and target system until the
aspects that require business sign-offs and
migration is executed fully. Reconciliation
commitments. A RACI matrix can help delegate
is done at the end of each day until all the
the roles and responsibilities at an early stage
data is migrated.
of the project, thus setting expectations for all
the stakeholders. Environmental factors influence the choice
of migration. The source and target system
• Strategy planning: Early stage planning determine the conversion methodology and the
helps to identify the potential issues/risks data migration tools to be used. The extract,
that may occur later in the project, enabling transformation, load (ETL) tool tends to be
banks to plan for risk mitigation. One of the preferred over other technologies for its ability to
important aspects that needs to be addressed handle large and complex volumes of data.
and planned up-front is the rollout strategy.
As Figure 2 (next page) shows, there are three Release planning is critical to determine when
basic strategies to consider for a rollout plan: a given functionality will be delivered so that
>> Big Bang: Migration is done in one single sample ETL iteration can be planned before a
major release to ascertain any potential impact
operation. It is usually undertaken over
of the migration.
the weekend. This is preferred for low data
volumes.

cognizant 20-20 insights 3

Big Bang vs. Phased vs. Parallel

➤
BIG BANG PHASED APPROACH PARALLEL RUN

In this type of data migration In this type of data migration, In this type of data migration
the entire data is migrated in the data is moved in a phased the data is migrated in parallel
DESCRIPTION
one go. manner. along with business operations
module wise.

Suitable when the volume of Suitable when the volume of Suitable when the volume of
IDEAL data is low such as in the case data is large such as in the case data is large such as in the case
SCENARIOS of small banks. of large banks. of large banks.

• Short implementation period. • Less risk. • Business not affected.

• Lower cost. • Training on the job. • Switchover to new system
ADVANTAGES • Small volume of data to only when everything is
validate. functioning well.
• Less downtime.

• High risk. • Longer implementation time. • High cost.

• Small details can be • Disruptive at times. • All data exists twice.
DISADVANTAGES overlooked. • Gaps during transition phase. • Maintenance of existing systems.
• Failure in one part can have • No fallback after first phase is • Handling changing process
cascading effect. implemented. difficult.

Figure 2

Rollback strategies should be planned to regain system. This phase is a good opportunity to get
the original state of the system in situations rid of redundant/unwanted data. The following
where migration has been inadequate. The activities should be performed to ensure the
decision to roll back should be taken before the sanctity of the data:
target system goes live. A detailed rollback plan
consists of criteria for rollback, steps to roll back • Data profiling (source & target): Data profiling
the target and source systems to their prior state helps automate the identification of data and
and testing of the system in its rolled-back state. metadata while enabling the correction of
inconsistencies, redundancies and inaccura-
• Change management board: The change cies. Source and target data are profiled to
management board is responsible for initiatives discover data structure and relationships.
to establish an effective change management
process. The board must vet all the changes
• Product listing: Product listing is a process in
which each corresponding product in the new
required and schedule them for implementa-
system is mapped to the existing product in the
tion based on the analysis.
legacy system. In some cases, the products are
• Project schedule and initiation: Once the rationalized to fit the supported products in
size of the task is understood, a proper project the new target system.
governance structure needs to be imple-
mented. It has been observed that 85% of
• Data mapping: The data conversion process
begins with data mapping, which essential-
the migrations fail or experience delays.2 An
ly entails mapping the legacy system data
optimal project delivery structure helps plan elements to that of the target system. This
for sufficient contingency. The structure must process should ensure a comprehensive
include project timelines, deliverables and mapping between the source and the target
milestones. systems. All the data fields that are going to
Detailed Analysis be migrated must be examined in terms of
The analysis phase is probably the trickiest part the data types, field length, system-specific
of the process. The data can be analyzed well only rules and integrity checks. Data mapping is an
if it is understood well. Migration should be driven iterative process and for every change in the
based on the target system and not the source design or rule of the system, changes should
be incorporated in the mapping specification.

cognizant 20-20 insights 4

• Data cleansing: A data migration project’s main rectifying incorrect and inaccurate records
objective is to migrate clean data into the core from a data set. Some of the methods that
banking system. This requires cleansing legacy can be used to cleanse data are the manual
data, which can be very elaborate depending approach, automation or combinations of both.
on the method chosen and hence must be Training: End-user training for the new target
planned carefully. It helps form a strong base system before “go live” is important for a smooth
for the entire ETL process by detecting and transition and effective usage of the system.

Representative Examples of Product Listing/Data Analysis

LEGACY CORE DATA MAPPING

SYSTEM BANKING SYSTEM CONSIDERATION/SOLUTION

Customer Corresponding product in the core

System 1 record 1 banking system for the multiple
Customer Single Customer ID entries created in the legacy system
System 2 Customer has to be listed and the data mapped
record 2 to a single Customer ID.

BALANCES
Savings Account
Savings For effective migration of different
Type 1 Netted Savings Balance
Account portfolios, balances should be netted.
Savings Account
Type 2

Capture the outstanding balance of

ACCRUALS Account Deposit contracts and bills.
Settlement
Fixed Term Deposit Map the settlement account and
Interest Interest interest capture account against the
product.

Customer Accounts Customer Accounts

Accounts present in the legacy system
can be migrated individually.
Corporate Accounts Corporate Accounts In case of closed accounts, year-end
tax documents can be migrated.
Historical/Closed Tax Documents

Figure 3

Migration Configuration • Configure staging area: A separate section

• Configure migration tool: The data migration is allocated in the production server to store
tool acts as a central repository to support extracted files generated after transformation.
storage and maintenance of data mapping. The It is used to store the intermediate results of
tool selected for migration should be flexible the ETL process. The target tables should refer
and scalable, ensuring efficient ETL, and should precisely to the final database, ensuring faster
also be capable of an automated transfer. The migration to the target system.
tool should be customized according to the
• Execute data conversion: Once the staging
target system. area is configured, data from the legacy system
can be moved to the target system. The data
is converted from the structural form that it
exists in the legacy system to the required form
in the target system.

cognizant 20-20 insights 5

ETL Approach

EXTRACT TRANSFORM LOAD

Data is Data is Data is
extracted from transformed transported to
the legacy to a the core
system standard banking
schema system

Figure 4

Mock Runs & Testing • Load on to target system: Data uploaded to

• Trial data load: As a pilot effort, small samples the staging area is loaded in sequence to the
of data are uploaded to ensure load sequence, target system using the migration tool and
determine duration and accuracy, and correct data load scripts.
any loading errors. This is iterated till the load
• Verify target application: Once the data
process completes successfully. has been loaded on to the target system, it
• Mock run: This is a mockup of the entire ETL is verified for accuracy by performing recon-
process to check for any load errors during the ciliation checks that could be in the form of a
process. In case of any error, it is reiterated report. If any discrepancy is identified, then the
with corrections till the process completes suc- systems are rolled back to their original state.
cessfully. If no error is found, then the team can
• Target system implementation: At this stage,
go for final migration. the entire ETL process is completed and
• Testing: An effective testing strategy is critical the target system is ready to go live. All the
to reduce the risk associated with a data required business rules and acceptance criteria
migration process. Testing can be performed in are verified and key performance metrics are
the following phases: evaluated.

>> Unit testing: This includes verifying the Reconcile & Go-Live
scope, data mapping, target system re- This is the post-migration stage where the target
quirements and migration tool used. Also, system is ready for use.
each unit will be tested as part of the func-
During the cutover period, the source system
tional end-to-end strategy.
has to be brought to a logical accounting stage.
>> Post-migration testing: This is done once During the migration process, no new transac-
the migration is executed. It includes test- tions must be entered into the legacy system and
ing the throughput of the migration process data must be frozen. Once the data migration
and reconciliation. is executed, reconciliation checks need to be
>> UAT: The functional test on migrated data performed to ensure there are no mismatches
in the target system is validated as per the before going live. Once the target system goes
requirement specifications. live, it has to be monitored to gauge success and
note any improvements required.
Migrate
This is the final stage of migration into the target Migration is usually executed over a weekend, and
system. A cutover period is defined for the entire all the stakeholders are informed of the planned
migration process to be executed. This is usually downtime. The channels interfaced to the system
the period between the shutting down of the will not be functional during this period. If the
legacy system and the new banking system going bank is required to support any immediate
live. Data in the legacy system is frozen and then transactions, they can be carried out in an offline
extracted during this period. The process outlined mode by creating records that are fed in to the
below is followed for successful migration: system later.

cognizant 20-20 insights 6

Best Practices and Processes Recommendations
The data migration methodology described above A seamless core banking transformation can be
provides an end-to-end solution covering the achieved by using proven data migration method-
different stages involved in the migration process. ology services. Successful migration adheres to
the following principles:
Some of the best practices that should be incor-
porated in this process are: • Banks undergoing transformation should have
clear business directions and goals.
• Ensure consistent alignment with the project
delivery team. • The program roadmap must be jointly
developed by the business and IT leadership.
• Set up timely meetings with the high-level
decision-making body consisting of bank stake- • Deploy a proven methodology to minimize
holders to monitor progress and address any disruption and mitigate risk.
issues. • Embrace effective change management.
• Cleanse data before migration and ensure • Facilitate product rationalization and process
there is enough knowledge about the data to restructuring for optimal utilization of the new
aggregate and maintain control over it. system.

• Create a reusable blueprint of the migration Conclusion

methodology by using previously created tools In today’s dynamic environment, banks tend to
and run books. Document the process in order consolidate and become global banks by imple-
to leverage previous experiences and improve menting a standard core banking system. An
on them. effective migration methodology enables banks
• Perform mock runs before the actual exercise to upgrade to the latest system with minimal
to ensure success and insulate from risks. business disruption. It also reduces the risk
involved in managing the complex process. With
• Plan for contingencies and expect time and
strong program governance, banks are equipped
cost overruns because of its complex nature.
to account for most possible errors and malfunc-
• Manage the cutover period strictly so that the tions by bringing in the right people, implement-
activity does not run beyond the period. ing global best practices and partnering with the
• Quickly fix any glitches to minimize the right vendors.
turnaround time.

• Migration rollback should be planned in case of

failure in migration or data reconciliation.

Footnotes
1 assets1.csc.com/big_data/downloads/DMI_For_SAP_Banking_Flyer.pdf
2 www.dmnews.com/dataanalytics/data-migrations-challenges/article/412805/

References
• www-05.ibm.com/cz/businesstalks/pdf/Core_Banking_Modernization_Point_of_View.PDF
• www.banktech.com/core-systems/12-best-practices-for-a-core-banking-upgrade/d/d-id/1294650?
• www.oracle.com/us/industries/financial-services/idc-ap772404u-wp-488545.pdf
• www.oracle.com/technetwork/middleware/oedq/successful-data-migration-wp-1555708.pdf
• www-935.ibm.com/services/multimedia/The_value_of_transforming_core_banking_systems.pdf
• www.mckinsey.com/insights/business_technology/overhauling_banks_it_systems
• support.sas.com/rnd/migration/planning/validation/parallel.html

cognizant 20-20 insights 7

About the Authors
Rajesh Ramachandran is a Product Consultant within Cognizant’s Banking and Financial Services’ Product
Solutions Practice. Rajesh has more than 11 years of experience in leading business and IT engagements,
mainly in the banking domains. He can be reached at [email protected].

Ravishankar Natarajan is a Product Consultant within Cognizant’s Banking and Financial Services’ Product
Solutions Practice. He has more than 12 years of experience as a consultant implementing various core
banking products. Ravi’s areas of expertise include retail banking, private banking, credit risk and data
migration. He can be reached at [email protected].

Senil Abraham is a Business Analyst within Cognizant’s Banking and Financial Services’ Product Solutions
Practice. He has more than three years of experience in IT engagements, mainly in the retail banking and
insurance domains. Senil can be reached at [email protected].

Rajdeep Bhaduri is a Lead Product Consultant within Cognizant’s Banking and Financial Services’
Product Solutions Practice. He has more than 17 years of experience in leading business and IT
engagements, mainly in the private banking and capital markets domains. He can be reached at
[email protected].

About Cognizant
Cognizant (NASDAQ: CTSH) is a leading provider of information technology, consulting, and business
process services, dedicated to helping the world’s leading companies build stronger businesses.
Headquartered in Teaneck, New Jersey (U.S.), Cognizant combines a passion for client satisfaction,
technology innovation, deep industry and business process expertise, and a global, collaborative work-
force that embodies the future of work. With over 100 development and delivery centers worldwide and
approximately 255,800 employees as of September 30, 2016, Cognizant is a member of the
NASDAQ-100, the S&P 500, the Forbes Global 2000, and the Fortune 500 and is ranked among the top
performing and fastest growing companies in the world. Visit us online at www.cognizant.com or follow
us on Twitter: Cognizant.

World Headquarters European Headquarters India Operations Headquarters

500 Frank W. Burr Blvd. 1 Kingdom Street #5/535, Old Mahabalipuram Road
Teaneck, NJ 07666 USA Paddington Central Okkiyam Pettai, Thoraipakkam
Phone: +1 201 801 0233 London W2 6BD Chennai, 600 096 India
Fax: +1 201 801 0243 Phone: +44 (0) 20 7297 7600 Phone: +91 (0) 44 4209 6000
Toll Free: +1 888 937 3277 Fax: +44 (0) 20 7121 0102 Fax: +91 (0) 44 4209 6060
Email: [email protected] Email: [email protected] Email: [email protected]

­­© Copyright 2016, Cognizant. All rights reserved. No part of this document may be reproduced, stored in a retrieval system, transmitted in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise, without the express written permission from Cognizant. The information contained herein is
subject to change without notice. All other trademarks mentioned herein are the property of their respective owners.
Codex 2287