Big data and advanced analytics are at the center of how financial services institutions are equipping themselves to deliver better value to their customers, while decreasing operating costs and mitigating credit, market, and operational risks.
By Dr Mark Nasila, head of advanced analytics: consumer and retail at FNB.
Big data is a term used to describe the exponential growth and availability of data, which may be structured, semi-structured or unstructured. These extremely large data sets may be analysed computationally to reveal patterns, trends, and any possible associations. Data science is an interdisciplinary field looking at processes and systems to extract knowledge or insights from data in various forms while advanced analytics is a set of analytical techniques used to predict future outcomes of several events.
Making use of data and analytical techniques has proven to help organisations meet regulatory and reporting requirements, compete with technology disruptors, and reduce operating costs. A large number of financial institutions have embarked on a journey of combining and making use of the available internal datasets, such as debit and credit transactions, transfers, channel and communication preferences, rewards utilization and loyalty behavior, etc. These institutions including banks have traditionally collected vast amounts of data from traditional sources such as transaction details and spending behaviors.
There has also been emergence of non-traditional datasets that can also be mined to extract meaningful insights. These include datasets from newer or external sources; website logs, Internet clickstreams, social media activity and mobile-phone call details. They feed into various internal or external systems at a growing rate.
The growth is attributed to advances in social media, which has also led to the generation of perception data. These types of datasets can equip banks with insights to effectively understand customer needs in several ways. These include developing algorithms around product or service sentiment data (e.g. Facebook, Twitter feeds) that can enable banks understand how customers perceive their products or services, analyzing response trends on the launch of new products as well as analyzing unstructured voice recordings from call centers to be able to recommend solutions to reduce customer churn, up-sell and cross-sell products and proactively detect fraud.
For FNB, the mining of big data provides a golden opportunity to continue being much better in several banking aspects. By using data science and advanced analytics techniques to collect and analyze big data, we can reinvent nearly every aspect of banking. These techniques enable targeted marketing, optimised transaction processing, personalized wealth management advice, prevention of internal and external fraud, assessing regulatory risks and much more – the opportunities are endless. A large proportion of the current big data projects in banking revolve around customers – driving sales, strategising retention, improving service and identifying needs: so the right offers can be served up at the right time to the right customer.
Some of our sales are already taking place based on automated, pre-approved offers on digital platforms based on the predicted propensity of the customer to take the products up. We are now able to model our customer’s financial performance on multiple data sources and scenarios. Big data and analytical techniques play a salient role in strengthening risk management in areas such as card fraud detection, financial crime compliance, credit scoring, stress-testing and cyber analytics.
The main objective of data science and advanced analytics is to obtain insights and knowledge from data to drive strategic decision making. These techniques require a methodical exploration of data using data mining techniques, enhanced by proven scientific techniques. Data science and advanced analytics combine elements, techniques and theories from many fields, including mathematics, machine learning, computer programming, statistics, data engineering, pattern recognition and learning, visualization, actuarial science, uncertainty modeling, data warehousing and high-performance computing.
Experts in these fields are responsible for examining the data, identifying key trends, and writing the complex algorithms that will see the raw data transformed into a piece of analysis or insight that the Bank can use to gain a competitive advantage. The advanced data science techniques are selectively and smartly applied to enable the bank to understand the correlation between the events being predicted and the data dependent factors driving the outcome of these events.
Historically, banks have been efficient at running analytics at a product level, such as on Cheque accounts, Overdrafts or Credit cards. With big data and analytical techniques, we are now able to get a holistic view of a customer across their portfolio. This is by consolidating different sources of data to obtain a single view of the customer, hence we can explore across all inter-connected customer relationships the customer has with the bank.
We can also understand customer relationships by exploring their ever-changing transactional and interactional behaviors. This will also enable us to effectively target customers and improve engagement as we interact and communicate effectively across new digital marketing platforms such as Web sites, e-mail, mobile apps and social networks. This is because advanced segmentation strategies on big data are enabling the bank to enhance its marketing strategies by identifying customer needs based on their transactional behaviors. For example, when a customer begins a relationship with us, we are well equipped to offer the right products at the right time, and even quickly resolve their complaints or queries by analyzing how they interact with various platforms.
Moreover, we are now in a position to determine an estimated cost for how each customer interacts with each channel (e.g., call center, branch banking, etc.) and come up with a recommendation of how we can move customers to low-cost channels. In the debt collections and recoveries space, analytics plays a salient role in enabling us create an accurate picture of a customer’s willingness and ability to pay and, hence estimate the amount likely to be recovered. This provides a criterion for prioritizing collections activities to maximize recoveries and reduce collections costs.
In the olden days, financial institutions were impeded by the difference in the time data is collected, the time data is analysed and the time decisions are made based on the analytics insights. This challenge is mainly attributed to legacy platforms and databases sitting in silos that were not specifically designed for the present data science needs. FNB has made a monumental stride in acquiring systems such as Netezza to provide capability for real-time or almost real-time analytics.
These capabilities provide us with new ways to innovatively solve business problems such as; proactively fight Financial Fraud: we can now timely analyze transactions, account balances, spending patterns, credit history, employment details, location etc. to determine whether transactions are legitimate. If unusual activity is detected, we can immediately action preventative measures, as well as alert the owner. We are now able to improve on our risk rating models i.e. a continuous feed of internal and external data implies credit, regulatory and market risk ratings can be updated in real time. This enables us to quantify these risks more accurately.
Moreover, we as a bank are now able to have a more accurate picture of a customer’s assets, business operations and transaction history at a specific time of their lifecycle. Furthermore, we can now determine the value of our customers more accurately: This entails using this continuous feeds of data to determine the future behavior of customers, and tailor policies to account for a customer’s changing financial situation at a particular time of their lifecycle.
For us to gain and sustain a competitive edge in the big data and advanced data science space, we need to continue to actively identify components of the big data trends that are a right fit for advancing our businesses strategy. Not all data, but relevant data coupled with innovative advanced data science techniques will prove transformational in maintaining us as a powerhouse in the financial industry.