Even when there were no computers, banks had put in place an efficient system of recording various transactions. Most business transactions took place at branches, which were supplying both management and regulatory reports. These reports were manually consolidated at intermediate controlling offices for eventual aggregation at the corporate level. These manual systems worked well till the scale of operations were relatively small.

As the banks grew in size and expanded geographically, the volume of transactions became quite large. Manual aggregation became both time consuming and error prone. Banks with a large number of branches spread across geographies, began using computers to automate the aggregation process. Despite these efforts, management information system (MIS) in the banks had the following drawbacks.

• Time lag (aggregation held up till each branch has reported)
• Data quality (each stage of consolidation and aggregation was a source oferrors)
• Unavailability of customer specific data (customer identity shrouded byproduct centric record
• keeping and branch data encapsulation in batchprocessing)
• Data granularity required for developing analytics (what if scenario, drill down) was not available to decision makers.
• Reporting activity competed with business activity for resources at the branch.
• Data classification rules were not applied uniformly across the organisation, and also varied with time.

In India, majority of banks began using information technology for MIS. The inflexibility of Cobol programmes and batch processing was soon overcome by powerful desktop systems with rudimentary database systems, which allowed banks to analyse data, once it has been received in manual form from branches, transcribed into machine readable formats and validated. Quite a few of regulatory (which banks termed as statutory in those times) reports were also produced in this way. These earlier initiatives laid the foundations of BI in banking.