 |
Multiple copies of the same data have always
plagued the agility of institutions, in general, and
financial institutions in particular. For example, I
have two logins from my bank, one to access my
banking accounts and another to access my credit
card information. If at a later date my residence
address changes, I will need to login to two
systems and change the address, in both instances.
This problem arises because there are two separate
master data copies maintained, in this case for
customer data.
There are multiple definitions of enterprise master data.
Based on our experience in dealing with several banks,
we define enterprise master data as:
- Information used by multiple systems within a
bank. For example, customer data shared
by multiple systems within a bank. Data specific
to an individual system cannot be considered
enterprise master data. A branch front-end system
may maintain teller-wise limits to post transactions
through the branch. Assuming that this processing
is carried out completely within the branch
front-end, this information would be master data
specific to the branch front-end system and not
enterprise level master data.
- Master data is required to perform transactions
but is distinct from transactional data. Master data
represents nouns and transactional data represents
actions. For example, currency conversion rates
could be considered master data whereas foreign
exchange operations performed using this data
would be considered transaction data.
|
