Data Warehousing To Drive Innovation At Bank Of Baroda

The Bank of Baroda (BoB) is in the process of implementing the HP Neoview data warehouse solution. The bank initiated the deployment in June 2008 and will have the first decision support information report by January 2009. This report will help to expedite the implementation process.

DW provides a platform for innovation

“Data warehouse is a core application for any enterprise that can be used as a framework for developing a number of other arterial solutions around the warehouse like MIS, BI, Performance Analysis etc,” says K M Asawa, DGM – Projects and IT – Infrastructure, Bank of Baroda. Basically, the data warehouse solution provides a predictive model. The solution acts as a pool of company information that is extracted from various sources like the Core Banking Solution (CBS), the Human Resource Management System (HRMS) and other applications. After pooling the information, it is analysed, processed and aggregated into small silos as data models.

Every organisation creates independent data models. For example, a bank will have to develop a model to aggregate the total deposit balances for an entire year to calculate the monthly average of deposits received by the bank for that year. Each business unit will have to develop hundreds of data models to arrive at various conclusions.

Potential implementation hurdles

“There is a point of caution that must be kept in mind before one gets carried away with the immense potential that a data warehouse provides simply because Data Warehouse implementations traditionally have a very low success rate,” says Asawa. Reasons for this could be that DW solutions have huge manpower demands and necessitate the building of data models with long gestation periods. Also, any organisation needs to have records of at least five years of production data for developing these data models, which would then predict the right trends.

The other challenge is the quality of data. Most PSU banks do not have updated data, especially in areas like the demographic information of customers, which includes details like postal address, phone number etc. If the information is dated, the models will not provide accurate predictions or analysis. The other side to the data quality issue is the challenge with the legacy systems. They fail to provide consolidated information about the customer. A comprehensive customer profile can help the bank to devise appropriate marketing plans for different customer segments.

Another major hindrance to a successful Data Warehouse implementation could arise from a lack of vision to think about the market scenario three or five years down the line. Lack of vision could also entail lack of clarity about what the organisation expects from the Data Warehouse implementation. “It is very important for the bank to decide what it wants from the Data Warehouse. Subsequently it can decide on data models, which can further be put to use by analysts or for that matter any decision making body of the bank.”

Role of Top Management

The most important challenge to overcome is the lack of vision. The Data Warehouse exercise at the BoB has been designed to involve top functionaries at the bank as they will be best suited to give insights about the organisation, industry trends, RBI guidelines etc.

To ensure data quality, the bank will either provide incentives to its staff to collect accurate data or hire a third party agency for the task. There are approximately thirty million customer accounts with the bank. BoB will employ a strategy that will capture the information on a priority basis starting with customers having sizeable deposits with the bank. This exercise will be done at the branch level. Each branch will define its own criteria for prioritising customers. Approximately 60 percent of customer related data will be captured initially.

Implementation Strategy

The bank intends to generate various MIS models and decision support information reports. The first report is likely to come be out by either January or February 2009.

This stage is known as the business discovery stage where the implementation team will discuss the business and MIS requirements with various business units like credit, deposits, cards, risk management, treasury and the international units. The response will be captured in a structured format. The information will be analysed and used for data modelling.

The next stage is the information discovery stage, which includes identification of information processes, listing them out, creating a data dictionary and mapping them to the right information fields once all the data is pooled.

These initial stages ultimately lead to the final products like CRM, strategic planning, predictive analysis etc. The project team includes people from HP and the bank staff.

Benefits

Data Warehouse enables better and improved decision-making by looking at deposit performance analysis and loan performance analysis. The marketing department also gets better and focussed content about customer behaviour patterns and trends. Data warehouse also allows historical v/s current trend analysis to facilitate strategic decision making. The solution helps banks in risk management and compliance.

“Compliance is one of the most critical areas where local banks need to cover ground if they have to match the foreign banks after the sector is liberalised next year (by March 2009), which will allow foreign banks to buy stake in Indian private banks,” says Asawa. Apart from other competitive parameters, technology is also a parameter that will determine the success rate of banks.

On a concluding note, a Data Warehouse solution provides product analysis and encourages innovation, which acts as a profit-generating avenue for the organisation.