Harnessing Big Data in Banking
- Anjani Kumar
Proactive banks understand that Big Data can be harnessed for risk-based, real-time pricing, unified customer view, product differentiation, compliance and risk management, fraud detection and prevention of false positives, product and service development, customer segmentation and targeting, customer retention and loyalty enhancement, cross-selling and optimized offers, and much more.
Many banks have implemented Big Data solutions and are leveraging technologies like Hadoop for integrating heterogeneous reference data sources and distributing their data across the bank in real-time. Others are using Big Data solutions for globally integrating assorted banking solutions for better business decisions.
But away from the success stories, a large number of banks are still struggling with the why and how of Big Data, unable to capitalize on its opportunities.
Here are a few ideas on what banks should do to improve the effectiveness of their Big Data solutions:
1. Strategic Planning: Banks must include Big Data in overall strategic planning. Instead of focusing only on the internal data in a few business areas, they must consider Big Data holistically, including data enabling a single customer view, as well as that related to product, service, regulatory compliance and risk. It is also necessary to focus on external data - not just credit scores or market data feeds but also data from social and streaming media and more. In the early stages of Big Data implementation, banks could fully leverage in-house transactional data before turning to external data sources. Technology teams must identify and prioritize the areas of high business impact, such as customer-facing processes like sales generation and lead enhancement, to be targeted first. Banks must evaluate off-the-shelf solutions to see if they suffice or whether a deeper bespoke solution is needed. While Big Data implementation could start as a small standalone piece, it is important to integrate it with existing systems and applications. Maintaining balance between cost and function, and technical requirements and privacy considerations is crucial. As far as possible, only one copy of the data should be maintained to ensure reliability.
2. Robust governance and operating model. The Big Data and Analytics operating model and governance policies should be clearly defined. It is important to define how analytics would be embedded into the business, and the roles and responsibilities of all concerned. An executive champion must be empowered to enforce data discipline and governance across the organization. Drivers, objectives and success metrics must be clearly defined. Senior leaders must help to clear stumbling blocks. Predictive modeling must be used to run 'what if' scenarios and their associated cost/benefit analyses. A Big Data and Analytics innovation lab could facilitate quick idea generation and experimentation.
3. Information architecture. Banks must evaluate the robustness of their information architecture to ensure it can support the increasing complexity, volume and velocity of data. Most banks have complex base information architecture, spanning numerous products across myriad lines of business, geographies and channels. The new information architecture should enable an integrated, detailed view of enterprise-wide data and relevant external data, and facilitate data consistency, accuracy and auditability; it must also be agile, flexible and extendable. Formal data controls and governance will protect data integrity. To attain enterprise-wide data integration, banks should define their enterprise data architecture and roadmap, enable cross-functional data integration projects, enforce measurement of data quality, and institute processes for addressing data quality issues.
4. Privacy and security. Banks must leverage Big Data approaches to supplement fraud and risk management systems to bolster security and privacy. This will improve customer confidence and experience and also enhance the transparency of security processes.
Besides implementing the above, banks can learn from success stories like the following:
IBM has been enabling Mexico's Banorte Bank map out a new banking model and is using Big Data, marketing automation and innovations in analytics to create more personalized interactions and keener insights into consumer banking behavior. The Bank has redesigned its systems to advise staff on the products best suited to individual customer needs.
In 2014, Bank of North Carolina showed the way for community banks by enhancing investment in Big Data visualization. Using SAS Visual Analytics it standardized reporting, improved validation and report control and also enhanced speed and usability. The Bank can aggregate data better for portfolio reporting, as well as implement detailed reporting across personnel and business lines