Geoffery Nairn is an accomplished journalist and B2B copywriter with more than two decades of experience writing about IT, telecoms and the disruptive technologies shaping tomorrow’s world. His work has appeared in leading business publications such as the Wall Street Journal, the Economist Intelligence Unit and the Times, and he was a core contributor to the Financial Times for more than a decade.
Leveraging this extensive experience as a technology journalist, he now works primarily for commercial clients in the technology sector, specialising in high-impact projects designed to catalyse change, deliver action or reinforce thought leadership.
Against a backdrop of record low interest rates, commoditisation and fierce disruptive competition, banks desperately seek a new growth agenda. They need to attract new customers and improve the profitability of their existing customer base.
To do that, banks have to improve the customer experience by leveraging the power of their data using advanced predictive analytics tools to help build better relationships.
Banks have been having a tough time in recent years. The financial crisis created huge holes in their balance sheets and left customers with a profound distrust of financial institutions that will take time to completely erase.
At the same time, the technology revolution of recent years has revolutionised the concept of customer service. The internet, smartphones, and social media have brought new ways for banks to interact with their customers, who expect banks to offer them a much more interactive and personalised customer experience.
They expect banks to be as responsive and “user-friendly” as a mobile app. Startup banks can do this as they can design their products, business processes and IT systems to be much simpler and customer-focused.
But traditional banks are weighed down by complex product sets, legacy operating models and siloed information systems that make it difficult for them to be competitive in areas such as product innovation and customer service.
Banks increasingly recognise the need to rebuild their organisations around the customer. Two out of three bank executives believe that a customer-centric business model is ‘very important’, according to the Retail Banking 2020 report by consultancy PwC.
But the big challenge today is that there is a significant gap between awareness and preparedness. Only 17 percent of the executives surveyed by PwC felt they were ‘very prepared’ to become customer-centric.
Traditional banks should be at a great advantage over new entrants when it comes to achieving customer-centricity, as they have a wealth of data on their customers that goes back many years.
But this data is fragmented across multiple legacy systems that are siloed by business function or product group, and the data may be stored in incompatible formats.
That makes it very difficult for a bank to ‘join the dots’ and get a complete picture of a customer, without which they will continue to struggle to understand what a particular customer really wants from their bank.
Recent years have seen huge advances in information and communication technologies that have transformed customers’ expectations of what a bank should be.
In the past, bank customers might check their balance once a week in an ATM. Today, mobile banking customers may check their balance twice a day and they expect the information to be updated in real-time.
The increase in transaction volumes created by internet and mobile banking places heavy loads on bank systems, but it also provides valuable information on how customers interact with their banks in a digital age.
To understand their customers better, banks collect transactional data and a wealth of other structured and unstructured information from traditional sources such as credit scores and customer surveys, as well as new non-traditional sources such as social media.
But as the amount of data managed by banks grows exponentially, it becomes increasingly difficult to manage.
According to a survey by Bloomberg Businessweek Research Services (BBRS),
60 percent of the bankers surveyed, said that the volume of data from customer transactions and other sources was overwhelming their systems.
Banks risk drowning in information while they remain starved of knowledge. With more data arriving each day, it is clear that banks urgently need a radical makeover of their information systems if they are to make better use of the growing mountain of customer data to achieve competitive advantage.
Predictive analytics is one of the key technologies to make that transformation happen. If banks can get better at analysing customer data, they can gain more accurate insights and move from simply reacting to
events to being able to predict future trends much better than they currently can.
In the BRBS survey, 59 percent of respondents rated predictive analytics as an “extremely valuable” technology.
Most raw data is too big for human consumption so Business Intelligence tools were developed to describe what the data means in a way humans understand.
Predictive analytics is the next state in the evolution of data analysis, from being merely descriptive to being predictive.
Using predictive analytics, banks can formulate and answer much more interesting questions that look into the future: Which customers will generate the most profit with the least effort?
A bank might want to see how much it can optimally spend on acquiring or retaining customers in a particular segment, and to predict how their profitability might change over time by assigning different probabilities to a future 1 or 1.5 percent rise in interest rates, for example.
Prescriptive analytics is a more powerful form of predictive analytics that recommends a particular course of action and tracks the outcome produced by that particular action.
One of the more popular uses of analytics is to predict which customers most likely to respond to a particular marketing campaign. Currently, banks lack a 360-degree view of their customers and so cannot target campaigns effectively.
For example, the bank’s credit risk department will know that a customer has been refused a loan, but the marketing function will not. That’s because the marketing department typically only has access to aggregate information such as demographic data to target its campaigns.
So, the customer gets sent a promotional leaflet for a car loan because they are in the right age group, even though they are a poor credit risk and will probably not be successful if they apply again for a loan.
Such behavior not only wastes resources but risks alienating customers, particularly in an era in which customers are used to receiving personalised on-target messages from firms like Amazon, Google and Facebook, which are skilled at using customer data to create rich and personalised experiences.
By using predictive analytics, a bank can identify the best customer segments for a campaign and replace a “one-size-fits-all” approach with individualised, highly relevant messages tailored to each customer’s profile. That should lead to higher response rates and reduce the chance that promotional campaigns get marked as “spam”.
Predictive analytics not only enables the bank to get closer to its customers but to also optimise costs.
For example, the rise of online and mobile banking means customers visit traditional bank branches less often. High-cost branch networks cannot survive in their traditional form, so banks are looking to significantly reduce branch numbers in the coming years and reinvent the remaining branches so they remain relevant.
But if banks are too aggressive in their branch closure plans, they risk losing customers. That’s because customer loyalty is often based on the convenience of a local branch. If their local branch closes, customers could move to a different bank that still has a branch in their neighborhood, or change to a low-cost online-only financial service provider with no branch overheads.
Analytics can help banks predict which branches can be closed with least negative impact or enable them to develop alternate, less traumatic strategies, such as reducing opening hours or personnel in those branches that get less traffic.
Key bank, a US regional bank, used analytics to optimize its branch network by closing some branches and changing staffing levels at others to better match the needs of customers as they varied through the day. The bank estimates the changes helped them save $35 m annually.
Human resource issues are another factor as effective analytics programs require staff with skills that are often scarce and expensive.
Finally, the benefits of analytics initiatives will only be visible If the financial Institution is prepared to change to embrace new ways of working and to create new end-to-end business processes designed to overcome the functional silos and legacy attitudes.
Despite the challenges, it is clear that predictive analytics will play a key role in helping shape the bank of tomorrow. Those banks prepared to embrace this powerful technology today, will be better placed to build a sustainable competitive advantage.
Customers change banks for a variety of reasons, not just for the convenience of the branch network. Predicting which customers are likely to switch to other banks and what factors might trigger that action is another powerful application of predictive analytics. If a bank knows that a customer may be about to leave in the near future, it has time to prepare personalised offers to retain the customer.
Fraud analysis and portfolio optimization modeling are other promising applications for predictive analytics in banking.
Despite the many areas where predictive analytics could be usefully employed In financial firms, penetration levels for these tools are relatively low, in single digits, due to difficulties in operationalising analytics initiatives in large banks, which are not only hindered by legacy IT systems, but also outdated ways of thinking.
One big challenge Is that analytics Initiatives require internal data sources to be integrated, which is very difficult if a bank’s systems are siloed and data is fragmented according to different needs.
To understand customer behavior, analytics also require large amounts of external data, much of it unstructured, which imposes fresh challenges for IT departments used to only handling structured internal data.