What is the most difficult task for a bank? Acquiring customers? Retaining customers? Growing share of customer wallet? No. Being customer centric is the most difficult task for any bank. But why is being customer centric so important? The answer is simple. A customer who is better engaged is easy to retain, easy to acquire and easy to do business with.
So, how can a bank be more customer centric? It is not as easy as you think. This is where the importance of big data analytics comes in. Banks need to exploit the big data of customers to derive value for their customers. The equation is simple. Big data + analytics = big opportunities.
For example, I use my credit card in a shop. My bank acquires data about me, like what I bought, how much I spent, when and where I spent it etc. Banks offer a number of services, and with their large customer base, billions of data gets piled up in their database. This huge amount of data is called big data. So, the challenge for any bank is to utilise this big data and make sense out of the customer information so they can be more customer centric.
However, many banks are unable to exploit big data for various reasons. Data about customers typically gets stored in silos and there is no single common server of database to get a 360-degree view of the customers. A classic example was when Deutsche Bank attempted to implement a big data project, and they faced problems in extracting data from legacy systems.
Also, the time associated with implementing a data analytics project is considerably huge. Nearly 75% of banks do not have sufficient talent for such analytics. More importantly, the lack of interest shown by top management is also one of the biggest challenges. They need to move on from their traditional approach and start thinking about new technology. Then there is the unstructured data problem that many banks are finding difficult to overcome.
There’s two types of data – structured and unstructured data. Structured data can be easily organised. For instance, any data that is entered into a computer like age, gender etc in a prescribed format is structured data. Unstructured data is one that does not follow a specified format for big data. These are very difficult to analyse and this constitutes about 90% of the total data. For instance, assume that you are not happy with your bank’s services and you want to give them feedback.
You prefer to do it via Facebook or Twitter (since many of us are active in social media) rather than through the feedback form on the bank’s website. You give your feedback via the bank’s official Facebook page, stating: “I am not happy with your credit card service. I am receiving offers that are always expired and irrelevant to me. Please look into it.” Now, this is entirely in the text format, which makes it very difficult to analyse and most of the time it gets dumped in the database. This is one of the main reasons why your feedback is not considered by your bank and in turn they are unable to provide you the best banking experience.
The conversion of unstructured data into meaningful data is very important. This can be done by new technologies like natural language processing, text mining, stochastic-based algorithm etc. The “raw” text, typically in the form of tweets, Facebook comments and emails, is analysed by converting the text into words or phrases. These words or phrases are categorised as “good” or “bad” to analyse the mood of the customer. For instance, the word “good” gets a “+1”, the word “bad” gets a “-1” and “neutral” gets “0”. Subsequently, all the unstructured data in the form of text is converted into structured numeric, which makes up the input for analytical algorithms and then it can be processed further easily.
But one should be very careful in implementing a big data project. There are a series of steps that need to be followed while implementing a project, as suggested by IBM. Firstly, in the Educate stage, focus should be on knowledge gathering and market observations. In the Explore stage, banks should develop strategies and roadmaps based on business needs and challenges. In the Engage stage, they should pilot big data initiatives to validate value and requirements. In the Execute stage, the deployment with continuous application of analytics is done. The best way would be by adopting a “Start small and add more complexity step-by-step” strategy. Rabobank adopted this strategy and learned valuable lessons. A controlled journey is always better.
Though big data analytics in retail banking has not matured yet, many banks have already adopted big data analytics and reaped benefits.
Bank of America adopted big data to provide consistent, appealing offers to well-defined customer segments
U.S Bank adopted it to improve lead conversion rate by over 100%
Rabobank adopted it to analyse criminal activity in ATMs and it was ranked No. 10 among the world’s safest banks by Global Finance
To sum up, most banks are struggling to be more customer centric. Understanding customer needs is the key to achieving this goal. Technology has developed to make use of this big data and make meaningful insights. Therefore, it is time banks shift focus to customer analytics, as it benefits both customers and banks.
The choice is very clear for banks now. They could either develop a big data culture, capabilities and technology in the organisation and grow in the market or live at the mercy of their customers. Let us hope they wake up and make the right choice.